4G vs 5G Industrial Router: Which Suits Your Plant in India

For PLC and SCADA telemetry (tag values, alarms, remote diagnostics) a 4G LTE industrial router with dual SIM failover is usually enough, because the data is small and LTE coverage is wider.

Choose 5G when you have a measured need for high bandwidth (several camera streams, large file transfers) or lower latency, and you have tested 5G signal at the exact panel location on more than one operator.

Whatever you buy for use in India, check its TEC MTCTE certification status.

Who this is for: plant engineers and IT/OT teams choosing a cellular router for remote monitoring of PLCs, SCADA nodes, pump stations, DG sets or solar sites.

What is dual SIM failover?

The router switches to a second operator’s SIM when the first link fails, keeping telemetry online.

Tell us what you need on site

Need this on your plant? Tell us your requirement

Can I run PLC control over a cellular link?

Use it for supervision and data. Keep control and safety logic local in the PLC.

Start with the data, not the radio

Estimate what you actually send. A worked example (illustrative):

  • 300 tags, each a 4 byte value plus about 50 bytes of protocol overhead, sent every 10 s
  • 300 × 54 bytes = 16.2 kB per update, 8,640 updates per day
  • About 140 MB per day, roughly 4.2 GB per month, before compression or report by exception

Even this generous case is a small, steady stream, well within LTE uplink capability. Report by exception (send only on change, as MQTT and many SCADA protocols support) can cut it further.

One 1080p stream, by contrast, typically needs several Mbit/s continuously (the YouTube encoding guide recommends 8 Mbit/s for 1080p at 30 fps; H.265 CCTV streams are often lower, so check the camera) (YouTube recommended upload encoding settings).

That is where bandwidth starts to matter.

Measure your own: most routers show data counters. Run a pilot for a week and look at the monthly total and peak.

What 4G gives you

  • Coverage: LTE networks in India are much wider than 5G, especially in industrial estates outside city centres. Check operator coverage maps, then test at site.
  • Bandwidth: LTE category (Cat 1, Cat 4, Cat 6 and above) sets peak rates; for telemetry even Cat 1 is often enough. Categories are defined by 3GPP in TS 36.306 (3GPP specifications).
  • Power and heat: 4G modules generally draw less than 5G modules, which helps in sealed panels and solar powered sites. Compare datasheet figures for the exact models.
  • Cost: 4G routers and plans are generally cheaper. We do not quote prices here.

What 5G adds

5G NR is specified by 3GPP from Release 15 onward (3GPP Release 15). For a plant router it can add:

  • Higher uplink and downlink throughput, useful for video, large log uploads or remote desktop to many HMIs
  • Lower air interface latency, which helps interactive remote work. End to end latency over public networks still depends on the core network and internet path
  • Release 17 introduced reduced capability (RedCap) devices for mid range IoT (3GPP Release 17); check operator support before relying on it

What it does not do: make a closed loop control link over a public network safe. Keep control in the PLC; use cellular for supervision and data.

Most 5G routers fall back to 4G when 5G is not available. Confirm the fallback bands too.

Coverage is the deciding factor

Plants are hard RF places: steel sheds, concrete, panels inside panels. Before you buy:

  1. Use a phone or a test router at the panel location, not at the gate.
  2. Record signal values (RSRP, RSRQ, SINR) for each operator, for 4G and 5G. These are defined in 3GPP TS 36.214 for LTE and TS 38.215 for NR.
  3. Test at different times of day.
  4. If signal is weak inside, plan an external antenna on the roof with low loss cable. Antenna gain does not fix a dead site.

Check which bands each operator uses in your area and that the router supports them. India’s spectrum assignments are published by DoT (Department of Telecommunications).

Does an external antenna help?

Yes, when signal inside the panel or shed is weak but usable outside. It cannot create coverage where none exists.

Choosing a router for remote monitoring?

Tell us your tag count, update rate and site location, and we will help you size the link. Email the Embedos team

When is 5G worth it for a plant?

When you have measured bandwidth or latency needs, such as multiple camera streams, and verified 5G coverage at the install point.

Is 4G enough for SCADA?

For tag data and alarms, usually yes. Estimate data per month and test signal at the panel.

Dual SIM failover

For telemetry, reliability usually matters more than speed. Dual SIM lets the router switch to a second operator when the first fails. Ask:

  • Is it dual SIM single standby (one active at a time, switch on failure) or dual modem (both active)? Single standby is common and cheaper; switching takes time.
  • What triggers failover: loss of registration only, or a failed ping to a set host? Ping based checks catch more real failures.
  • Does it fail back to the primary SIM, and when?
  • Does the VPN reconnect automatically after a switch?

Pair it with local buffering at the gateway or PLC so data during an outage is sent later, not lost.

Security basics

Whatever the radio, follow the same rules as any remote access to control systems:

  • No open ports from the internet to PLCs or HMIs. Use a VPN or an outbound only broker connection.
  • Private APN from the operator if available.
  • Change default passwords, keep firmware updated, turn off unused services.
  • Log remote sessions.

CISA’s primary mitigations for operational technology say to remove OT connections to the public internet and, where remote access is needed, use a private network or VPN with phishing-resistant multi-factor authentication (CISA, Primary Mitigations for OT). The IEC 62443 series is the reference for zones and conduits (ISA/IEC 62443).

India buyer check: TEC MTCTE

Under the Indian Telegraph (Amendment) Rules 2017, the Telecommunication Engineering Centre (TEC) runs the Mandatory Testing and Certification of Telecom Equipment (MTCTE) scheme.

The TEC notice dated 1 July 2024 (Annexure-I) lists Routers under Phase IV, mandatory from 1 August 2024, and IoT Gateway under Phase III, mandatory from 1 January 2024.

There is no separate cellular router category, so ask the seller which category their model is certified under (TEC MTCTE notice). Ask the seller for the MTCTE certificate number for the exact model and check it on the TEC portal (tec.gov.in).

Also ask whether a WPC Equipment Type Approval (ETA) applies to the radio in that model (DoT).

Side by side

Factor 4G LTE router 5G router
Typical fit Telemetry, alarms, remote diagnostics Video, large transfers, low latency remote work
Coverage in industrial areas Wider Growing; test at site
Power and heat Lower, generally Higher, generally; check datasheet
Cost Lower Higher
Fallback Check whether any 3G/2G fallback exists with your operator Falls back to 4G
Must check Bands, dual SIM, MTCTE Bands, measured 5G signal, MTCTE

If your need is small Modbus telemetry over 4G rather than a general purpose router, see our 4G Ethernet Modbus gateway with Wi-Fi page for its 4G SIM, Wi-Fi and Ethernet options. 

Do industrial routers need TEC certification in India?

Telecom equipment in notified categories needs MTCTE certification. Ask for the certificate for the exact model and verify on the TEC portal.

Next step

Send your site location, tag count and whether video is in scope.

We will help you pick the link and the test plan.
Email: team@embedos.io

Help Us Help You Leverage The Power Of Industrial Automation & Industrial IoT – IIoT.

Modbus Register Map: How Do You Read It and Decode 32-bit Values (ABCD, CDAB, BADC, DCBA)?

Direct answer: a Modbus register map tells you, for every value, its data table (and so its function code), address, data type, scale and unit. Send the address as a 0-based offset (holding register 40001 goes out as address 0), read each 32-bit value as two consecutive registers, then try the byte orders ABCD, CDAB, BADC and DCBA until the number lands in the expected range. Our free Modbus register interpreter shows all four side by side.

Who this is for: plant and commissioning engineers connecting an energy meter, VFD or sensor to a PLC, SCADA or gateway, with the vendor’s register map in one hand and strange numbers on screen.

What you’ll be able to do after reading:

  • Pick the right function code for each row of a map.
  • Convert a 4xxxx register number into the address your master sends.
  • Decode 16-bit, 32-bit and float values and find their byte order from one known reading.
  • Apply scaling and units, and commission a new device with a checklist.

What does a Modbus register map contain?

A register map binds a device’s internal values to Modbus addresses. The Modbus Application Protocol Specification V1.1b3 calls this mapping “totally vendor device specific”, so two meters measuring the same voltage can look nothing alike on the wire. A usable map gives, for each value:

  • Parameter name and unit, such as line voltage in V.
  • Data table or function code.
  • Address, as a 0-based offset or a 1-based number such as 40001.
  • Data type and number of registers.
  • Word order for 32-bit values, and any scale factor.
  • Read only or writable.

If a column is missing, ask the vendor before you start polling.

Embedos blogs embedos-blogs: PLC rack with I/O modules and labelled wiring in an industrial control panel, the kind of Modbus master that reads a register map

Which function code reads which data table?

The table follows section 4.3 and the function code chapters of the application protocol spec. The 0x, 1x, 3x and 4x prefixes come from the older Modicon Modbus Protocol Reference Guide PI-MBUS-300 Rev. J, which many manuals still follow.

Data table Object Access Read function code Write function code Legacy reference
Coils 1 bit Read and write 01 05 (single), 15 (multiple) 0xxxx
Discrete inputs 1 bit Read only 02 None 1xxxx
Input registers 16-bit word Read only 04 None 3xxxx
Holding registers 16-bit word Read and write 03 06 (single), 16 (multiple) 4xxxx

One 03 or 04 request reads up to 125 registers; 01 or 02 up to 2000 bits. The spec lets devices overlay tables, so use the function code the map names.

Why is my reading off by one register?

The spec numbers elements in the data model from 1 to n but addresses them in the request from 0 to 65535, so a value numbered X is sent as address X minus 1. Its own example reads registers 108 to 110 with a starting address of 00 6B hex, which is 107.

The Modicon guide says the same thing in 4xxxx language: holding register 40001 is sent as address 0000, and 40108 as 006B hex. The leading 4 is not sent at all, because function code 03 already says “holding register”.

  1. A five-digit map number starting with 3 or 4 is a register number: subtract 30001 or 40001 to get the offset.
  2. Check your master. Some drivers take the register number and subtract 1 themselves; others want the offset. The driver manual says which.
  3. Prove it with a value shown on the device display. If it only looks right after shifting one register, you are off by one.

On 32-bit values this is easy to miss: you join halves of two different values and get a number that is wrong but not absurd.

Which data types will you meet in a register map?

The protocol itself only defines bits and 16-bit registers. Anything wider is the vendor packing bytes across consecutive registers.

Type Registers Range or format Typical use
uint16 1 0 to 65535 Status, scaled values
int16 1 -32768 to 32767 (two’s complement) Temperatures, signed power
uint32 2 0 to 4294967295 Energy counters, run hours
int32 2 -2147483648 to 2147483647 Signed totals, position
float32 2 IEEE 754 single precision: 1 sign bit, 8 exponent bits, 23 fraction bits Voltage, current, flow, pressure

The float format is defined in IEEE 754-2019, which is a paid standard. For a free explanation, the Wikipedia article on single-precision floating point is good further reading.

Embedos blogs embedos-blogs: single-phase DIN rail energy meter with screw terminals and kWh display

What do ABCD, CDAB, BADC and DCBA mean?

Inside one register the spec is fixed: big-endian, so 0x1234 is sent as 0x12 then 0x34. It does not fix which register comes first when a value needs two, and vendors chose differently.

Call the bytes of the first register you read A and B, and the bytes of the second register C and D. The four orders then mean:

Order Byte sequence used Common name
ABCD Register 1 high, register 1 low, register 2 high, register 2 low Big-endian
CDAB Register 2 first, then register 1 Word swapped
BADC Register order kept, bytes swapped inside each register Byte swapped
DCBA Everything reversed Fully reversed (little-endian)

Drivers and manuals use their own labels, such as “swap words” or “low word first”. Match on the byte pattern, not the name.

Worked example: decoding two registers in all four orders

An illustrative case, not a specific device: a meter’s map lists line voltage as float32 in two holding registers. Your master reads 0xB333 (45875) then 0x4367 (17255), so A = B3, B = 33, C = 43, D = 67.

Order Bytes float32 value
ABCD B3 33 43 67 -4.174 × 10-8
CDAB 43 67 B3 33 231.7
BADC 33 B3 67 43 8.354 × 10-8
DCBA 67 43 33 B3 9.218 × 1023

Only CDAB gives a believable line voltage, so this device sends the low word first. Set your master to word swapped and confirm against the meter display.

By hand, 0x4367B333 has sign bit 0 and exponent 0x86 = 134; minus the bias of 127 gives 27 = 128. The fraction 0x67B333 / 223 = 0.8101562, so the value is 128 × 1.8101562 = 231.6999969, shown as 231.7. Float32 cannot hold 231.7 exactly, so round to the map’s resolution.

The same check in Python:

import struct
r1, r2 = 0xB333, 0x4367            # registers in the order read
abcd = struct.pack('>HH', r1, r2)  # B3 33 43 67
cdab = struct.pack('>HH', r2, r1)  # 43 67 B3 33
print(struct.unpack('>f', abcd)[0])  # -4.1737987288570366e-08
print(struct.unpack('>f', cdab)[0])  # 231.6999969482422

In our free Modbus register interpreter, paste 0xB333 0x4367 in the order read. It lists float32 for ABCD, CDAB, BADC and DCBA, plus integer readings. Its built-in example, 0x4126 and 0x6666, decodes to 10.4 in ABCD.

Test with a non-zero value: 0.0 is 00 00 00 00 in every order. And confirm the data type first: read as int32, the same registers give 1130869555 in CDAB order.

Tell us what you need on site

Need this on your plant? Tell us your requirement

How do scaling factors and units work?

Many devices send a scaled integer instead of a float, with a multiplier in the map.

  • uint16 with scale 0.1 V: raw 2317 means 231.7 V.
  • int16 with scale 0.1 °C: raw 0xFF38 is -200, so -20.0 °C. Read as uint16 it becomes 6533.6 °C: a sign error, not a hot panel.
  • uint32 energy counter: registers 0x0001 and 0x86A0 in ABCD order give 100000. If the map says 0.01 kWh per count, that is 1000.00 kWh.

Watch units as closely as scale: W against kW is a factor of 1000. Some maps keep the scale in a separate register that you must read too.

How do I check raw Modbus RTU frames before decoding?

The Modbus over Serial Line Specification and Implementation Guide V1.02 puts a CRC-16 at the end of every RTU frame, sent low byte first. Check captured frames with our CRC and checksum calculator: its CRC-16/MODBUS preset gives 0x4B37 for the test string 123456789, sent as 37 then 4B. If the CRC fails, fix wiring or serial settings before blaming the map. For how RTU and TCP frames differ, see our post on Modbus RTU vs Modbus TCP.

Commissioning checklist for a new Modbus device

Check How to confirm Done
Map matches the device Model and firmware on the map match the label ☐
Function code per value 03 holding, 04 input, 01 or 02 bits ☐
Addressing convention Read a value you can see on the display; no off-by-one ☐
Data type Width, sign, integer or float ☐
Word and byte order Decode a non-zero known value in all four orders ☐
Scale and unit Compare with the display at two different values ☐
Sign handling Check a negative value where the process allows ☐
Block reads 125 registers maximum; avoid unmapped addresses (exception 02, Illegal Data Address) ☐
Writes last Only after reads are proven, with the process safe ☐
Record settings Order, scale and addresses in the commissioning sheet ☐

For a wider plan that starts with read-only functions across a whole line, see our guide to retrofitting legacy machines with OPC UA or Modbus.

Questions engineers ask about Modbus register maps

Is register 40001 the same as address 0?

In the Modicon convention, yes. Holding register 40001 is sent as address 0 with function code 03. Check whether your master wants the register number or the offset.

Why does my Modbus float read as a huge or tiny number?

The word or byte order in your master does not match the device. Decode the two registers in all four orders and pick the one inside the expected range.

What is the difference between input registers and holding registers?

Both are 16-bit words. Input registers are read only (function code 04); holding registers can be read and written (03, 06, 16).

FAQ

Does Modbus define a float data type?

No. The spec defines bits and 16-bit registers. Floats are a vendor convention that packs an IEEE 754 single-precision value into two registers.

Which byte order is the Modbus standard?

The spec fixes big-endian order inside each register. ABCD extends the same rule to two registers, but the spec does not mandate which register comes first.

How many registers can I read in one request?

Up to 125 with function code 03 or 04.

Does the register interpreter send my data anywhere?

No. The tool page states that decoding runs in your browser and nothing is sent. Its calculation basis is listed on the tools standards and disclaimer page.

Next step

Have two registers that will not decode? Paste them into the free Modbus register interpreter and compare all four orders against the reading on the device display. Embedos Edge is a platform of products, solutions, dashboards, insights and services.

Next step

Whatever master you use, decoding starts with the vendor’s register map.
Email: team@embedos.io

Help Us Help You Leverage The Power Of Industrial Automation & Industrial IoT – IIoT.

4-20mA Signal Isolator: When You Need One for Ground Loops

You need a 4-20mA signal isolator when a ground loop exists, that is, when the signal circuit is connected to earth or a common at more than one point and those points sit at different potentials.

Typical signs are a steady offset, a reading that moves when a motor or heater switches, or a value that changes when you disconnect a second receiver.

An analog signal isolator breaks the copper path between the field side and the PLC or IoT gateway analog input while passing the signal across. If the loop has only one ground reference and reads steady, you usually do not need one.

Who this is for: instrument and electrical engineers, panel builders and integrators adding a PLC, data logger or IoT gateway to signals that already feed another system.

Tell us what you need on site

Need this on your plant? Tell us your requirement

What is a ground loop in a 4-20 mA loop?

Earth is not one perfect zero volt point across a plant. Acromag’s note on how to select a 4-20 mA isolator puts it plainly:

Earth ground at different locations is often at a different potential, so grounding a loop in more than one place puts a conductor between two potentials and drives a current.

A DC difference shows up as an offset in the reading; an AC difference shows up as noise. When the difference changes as equipment switches, the fault comes and goes, which is why ground loops are hard to troubleshoot.

A 4-20 mA loop is fairly tolerant of wire resistance and pickup, because current is the same everywhere in a series loop.

TI’s overview of 4-20 mA current loop transmitters covers this. But that tolerance ends when a second path lets current leak out of the loop or in from another common. A 0-10 V signal is more sensitive still, because any voltage between the two commons adds straight onto the reading.

How do I confirm a ground loop rather than a bad sensor?

Measure loop current with a meter in series at the transmitter and at the receiver. If they differ, current is leaving or entering the loop somewhere. Then disconnect the second receiver or lift a suspect earth connection temporarily (safely, with the plant’s permission) and see if the reading returns to normal.

Does a shielded cable fix it?

Shielding helps with capacitive pickup, but a shield earthed at both ends can itself create a ground loop. The common practice is to earth the shield at one end. Shielding does not fix a potential difference between two signal commons.

Common situations that need isolation

Situation Why it causes trouble Typical fix
One transmitter feeding a DCS and a new gateway Two receivers with commons tied to different supplies or earths Isolator or splitter so each receiver has its own loop
Shared 24 V supply between field devices and gateway, with a grounded sensor Supply negative and sensor earth form a second path Isolator on the input, or separate supplies
Sensor electrically bonded to the process (thermocouples, some pH, some vibration sensors) The sensor body earth differs from panel earth Input isolated transmitter or isolator
0-10 V signal from a VFD analogue output to a separate panel VFD common and panel common at different potentials 0-10V isolator, or convert to 4-20 mA and isolate
Long runs between buildings Large earth potential differences, surge risk Isolator with suitable isolation rating plus surge protection

TI’s note also makes the point that isolation is required when a sensor can be electrically connected to a potential relative to the loop supply ground, which is common with thermocouples because they touch the material they measure.

Can I just add a second 250 ohm resistor for the new receiver?

You can put two receivers in series in one loop only if the loop supply can cover both input burdens plus the cable drop at 20 mA, and only one point in the loop is tied to ground. If both receivers ground their inputs, fit an isolator.

Types of signal isolators

Acromag’s white paper on signal conditioning and isolation lists the main reasons for isolation (breaking ground loops and protecting equipment from surges) and the arrangements:

  • Input isolation: input floats from output and power. Used when bringing field signals into a grounded system.
  • Output isolation: output floats from input and power.
  • Three-way (full) isolation: input, output and power are separated. Use this when three differently grounded systems meet.
  • Power isolation: signal input and output share a common, isolated from the power common.

By power source, you will see:

  • Output loop powered (sinking output): takes power from the receiver’s loop supply. Low cost, low input burden.
  • Input loop powered (sourcing output): takes power from the input loop. Adds a large voltage drop on the input loop, so check the transmitter supply budget.
  • Externally powered (four-wire): separate 24 V DC power. Most flexible, handles 0-10 V and conversions.

Acromag’s How to Choose a 4-20mA Isolator walks through the voltage budget for each type.

What isolation voltage should I ask for?

Isolation specifications should carry two ratings, a continuous working voltage and a breakdown or maximum test voltage. Compare like with like when you read datasheets. Pick the continuous rating based on the potential differences your site can actually see, with margin.

embedos-blogs-din-rail-4-20ma-signal-isolator-ground-loop-embedos-panel: DIN rail clamp terminal block, the kind of panel where a 4-20mA signal isolator is fitted before a PLC or gateway

When to isolate before a PLC or IoT gateway

A common case is retrofit: a signal already goes to a PLC or DCS, and now someone wants the same value on a remote monitoring gateway. Before you wire a second input in parallel or series, ask:

  1. Who powers the loop today, and where is its negative referenced?
  2. Is the gateway analogue input referenced to its own supply negative?
  3. Is the gateway supply shared with anything else in the panel?
  4. Is the sensor body earthed to the process?

If two or more references appear, put an isolator (or an isolating splitter) in, so each receiver sits on its own loop. If you are designing fresh, a single reference point and one supply per panel often avoids the problem without extra hardware.

Is isolation needed for every gateway input?

No. If the transmitter is floating, the loop has a single reference at the gateway, and the gateway is the only receiver, a direct connection is normal. Isolate where the checks above show more than one reference, or where surge exposure is high.

Checklist

  • Measure loop current at both ends; any difference points to a leak path.
  • Note every connection to earth or common in the loop, including shields and sensor bodies.
  • Check whether supplies are shared between field devices and the receiver.
  • Choose isolator power type from your voltage budget.
  • Choose isolation arrangement: input, output or three-way.
  • For 0-10 V signals over distance, consider converting to 4-20 mA.
  • Compare continuous isolation ratings, not just test voltages.
  • Earth cable shields at one end unless the device maker states otherwise.

Where Embedos Edge fits

Embedos Edge (for example Edison or Maxwell) accepts 4-20 mA and 0-10 V DC analogue inputs, so it is often the second receiver in case of a retrofit. 

The usual clean arrangement is: existing transmitter, isolator or isolating splitter, one output to the PLC and one to Edge.

Edge then sends the value over MQTT or HTTPS, holds data with data logging and buffering if the network drops, shows device health and connection status remotely, and can raise threshold alerts by SMS, email or call alerts. also Webhooks.  

Next step

Where the source is a drive or meter with a digital port, reading it over RS-485 Modbus RTU or Modbus TCP into Embedos Edge avoids the analogue ground loop question altogether.

For Your Analog Input Requirements, Email: team@embedos.io

Help Us Help You Leverage The Power Of Industrial Automation & Industrial IoT – IIoT.

Modbus RTU vs Modbus TCP: Frames, Wiring, Speed and When to Use Each

Modbus RTU and Modbus TCP carry the same function codes and the same register model, defined in the Modbus Application Protocol specification. RTU sends compact binary frames over a serial line, usually RS-485, with one master, a slave address and a CRC-16. TCP wraps the same request in a 7-byte MBAP header and sends it over Ethernet to port 502, with no Modbus CRC. Pick by the physical network you have or are building.

Who this is for:

Automation, instrumentation and maintenance engineers in plants, and system integrators, choosing between the two for a device, panel or retrofit.

Embedos blogs photo of a PLC rack mounted in a control panel, with field wiring in ducts and terminals above and below

What you'll be able to do after reading:

Read both frame formats byte by byte, apply the RS-485 limits correctly, estimate polling time, understand ports and security, and choose RTU or TCP for meters, VFDs, PLCs and remote sites.

Modbus RTU vs Modbus TCP at a glance

Point Modbus RTU Modbus TCP
Physical layer Serial: RS-485 two-wire is most common, RS-232 for short point to point links Ethernet, TCP/IP
Who may talk One master at a time per bus; slaves only reply Client and server; several connections, up to the device’s limit
Addressing Slave address 1 to 247, 0 is broadcast IP address plus a 1-byte Unit Identifier
Framing Frames separated by at least 3.5 character times of silence 7-byte MBAP header with a length field
Error check CRC-16 in every frame No Modbus CRC; TCP/IP and the Ethernet CRC-32 handle it
Maximum frame 256 bytes (253-byte PDU + address + CRC) 260 bytes (253-byte PDU + 7-byte MBAP)
Devices and distance 32 devices per segment without a repeater (4 fewer if the line is polarised); up to 1000 m at 9600 baud with AWG26 or wider cable Set by Ethernet design, switches and cabling
Port Not applicable 502; Modbus/TCP Security over TLS uses 802

All figures above come from the Modbus Organization’s free specifications: the Modbus application protocol specification, the Modbus over serial line specification, the Modbus messaging on TCP/IP implementation guide and the Modbus/TCP Security protocol specification. Wikipedia’s Modbus article is useful further reading.

Frame by frame: RTU address and CRC-16 vs the 7-byte MBAP header

The same request in both formats: read two holding registers (function 03) starting at address 0 from device 1. The specification numbers registers from 1 but addresses them from 0 on the wire. In the common Modicon-style convention many manuals use, holding register 40001 is address 0.

Why do I get exception code 2?

Illegal Data Address. The start address or quantity is outside the device’s map, often because the 40001 offset was not removed.

Embedos blogs diagram of the same Modbus request in two frames. RTU, 8 bytes: 01 slave address, PDU 03 00 00 00 02 (function 03, start address 0, quantity 2), then CRC-16 C4 0B low byte first. TCP, 12 bytes: 7-byte MBAP header 00 01 transaction ID, 00 00 protocol ID, 00 06 length, 01 unit ID, then the same PDU with no CRC. The shared PDU is boxed across both frames.

Modbus RTU request

01 03 00 00 00 02 C4 0B
01     slave address
03     function: read holding registers
00 00  start address 0
00 02  quantity 2
C4 0B  CRC-16, low byte first

Modbus TCP request

00 01 00 00 00 06 01 03 00 00 00 02
00 01  transaction identifier
00 00  protocol identifier (0 = Modbus)
00 06  length: 6 bytes follow
01     unit identifier
03 00 00 00 02  the same PDU, no CRC

The PDU in the middle is identical; only the wrapper changes. Check a Modbus CRC-16 yourself: choose CRC-16/MODBUS and low byte first.

Reading the response: why do floats read wrong?

A 32-bit float needs two 16-bit registers. Modbus sends each register high byte first, but the Modbus application protocol specification does not say which register comes first, so master and device can disagree without any error. A pressure transmitter on a 0 to 50 bar range returns 01 03 04 41 26 66 66 A4 4E over RTU, or 00 01 00 00 00 07 01 03 04 41 26 66 66 over TCP. The two registers are 4126 and 6666 (hex):

Order Bytes Value
ABCD 41 26 66 66 10.4
CDAB (word swapped) 66 66 41 26 2.718 × 1023
BADC 26 41 66 66 6.710 × 10−16
DCBA 66 66 26 41 2.717 × 1023

Only ABCD falls inside 0 to 50 bar. These values match the free tool to decode Modbus float word order (ABCD, CDAB).

RTU is a protocol, RS-485 is the wire

Modbus RTU is a framing; RS-485 is the electrical layer it most often runs on. From the Modbus over serial line specification:

  • Addresses: 1 to 247 for slaves, 0 for broadcast, one master at a time.
  • Silent interval: frames are separated by at least 3.5 character times.
  • Devices: 32 devices are always allowed on a segment without a repeater; more only if the device documentation says so. Transceivers rated at a fraction of a unit load raise this. TI’s RS-485 design guide notes that 1/8 unit load parts allow up to 256 nodes on an unbiased bus, but failsafe bias resistors can add up to 20 unit loads, which cuts that to about 96. The Modbus specification likewise reduces the allowed count by 4 when the line is polarised.
  • Trunk length: up to 1000 m at a maximum of 9600 baud with AWG26 or wider cable, halved when 4-wire cable is used as a 2-wire system. Category 5 cable may be used up to 600 m.
  • Drops: derivation cables never more than 20 m. Faster transceivers need shorter stubs, according to the same TI design guide (SLLA272).
  • Termination: at both ends of the trunk only, a 150 ohm resistor, or 120 ohm in series with 1 nF where the line is polarised. Never on a drop, and none on RS-232.
  • Bias (polarisation): one pair of resistors for the whole bus, pull-up to 5 V on D1 and pull-down to common on D0, between 450 and 650 ohm, usually at the master. The specification reduces the allowed device count by 4 when bias is fitted, and 650 ohm allows more devices than 450 ohm.

Check your values with the free RS-485 termination and bias calculator.

Is Modbus RTU the same as RS-485?

No. RTU is the framing and protocol; RS-485 is the electrical layer it usually runs on. RTU can also run on RS-232.

Should I use 8N1 or 8E1?

The specification default is even parity. If you use no parity, it requires 2 stop bits, though many devices ship as 8N1. Match the device.

How many devices can be on a Modbus RTU network?

Addresses run from 1 to 247, but an RS-485 segment is allowed 32 devices without a repeater unless the device documentation permits more, and 4 fewer if the line is polarised.

Speed and polling in practice: which is faster?

On RTU, the arithmetic

At the specification default of 19200 baud, 8 data bits, even parity and 1 stop bit, each byte takes 11 bits, about 0.573 ms. Reading 10 registers from one device:

  1. Request, 8 bytes: about 4.58 ms.
  2. Silent gap of 3.5 characters: about 2.01 ms.
  3. Response, 25 bytes (address, function, byte count, 20 data bytes, CRC): about 14.32 ms.
  4. Another 3.5 character gap: about 2.01 ms.
  5. Total wire time: about 22.9 ms, plus the device’s own response time from its manual.

With 10 such devices on one segment, a full scan takes at least 229 ms plus ten device response times. Above 19200 baud the serial line specification recommends fixed timer values: 1.75 ms for the inter-frame gap (t3.5) and 750 µs for the inter-character time-out (t1.5). Read registers in blocks, up to 125 per request with function 03. If a slave answers only some of the time, rule out clock baud rate error at either end.

On TCP

Wire time is rarely the limit. The device’s response time, how many simultaneous requests it serves, and any gateway in the path, which brings back the serial timing above, decide the real speed. “Faster” depends on the network and the devices, not on the protocol name.

Which is faster, Modbus RTU or Modbus TCP?

The network under TCP is faster, but a slow device answers slowly on either. Gains come from parallel connections, not from the single request.

Ports and security: 502, 802 and authentication

  • Modbus TCP uses registered port 502, as the TCP/IP implementation guide states.
  • Plain Modbus defines no authentication: any client that reaches the port can read and write. NIST SP 800-82 Rev. 3 notes that many OT protocols have no authentication at any level.
  • The Modbus/TCP Security protocol wraps Modbus in TLS, with X.509v3 certificates for authentication and role-based authorisation, on port 802. Check that both ends support it before relying on it.
  • Keep plain Modbus TCP on a segmented OT network, never on the internet.
What port does Modbus TCP use?

Port 502. Modbus/TCP Security, which wraps Modbus in TLS, uses port 802.

Is Modbus TCP secure?

Plain Modbus TCP has no authentication or encryption. Use network segmentation, and Modbus/TCP Security where both ends support it.

Multiple clients on Modbus TCP vs one master on an RTU bus

An RTU bus has exactly one master, and slaves never talk to each other. Modbus TCP uses a client and server model: the TCP implementation guide lets a server accept several connections and serve several requests at once, up to a limit set by the device’s design, and several transactions can share one connection using the transaction identifier. So two SCADA systems can read one TCP device directly, if the device allows enough connections, but two masters cannot share one RS-485 segment.

embedos-blogs comparison diagram: on the left, a Modbus RTU RS-485 bus with one master, three slaves on short drops and termination at both ends; on the right, two SCADA clients connected through an Ethernet switch to one Modbus TCP server, which accepts several connections up to its limit.
Can Modbus TCP have multiple masters?

Several clients can connect to one server, up to the device’s connection limit. An RTU bus allows only one master.

Modbus RTU over TCP is not Modbus TCP

Some serial device servers forward raw RTU frames, CRC and all, inside a TCP stream, often called “RTU over TCP”. That is not the Modbus TCP format: there is no MBAP header, and framing still depends on the CRC. A Modbus TCP client talking to such a server, or the reverse, gets garbage or timeouts. Set both ends to the same mode.

Converting RTU to TCP: gateways and the Unit Identifier

A Modbus TCP to RTU gateway is a TCP server on the Ethernet side and the single RTU master on the serial side. The TCP implementation guide says the Unit Identifier carries the serial slave address through the gateway; for a device directly on Ethernet it recommends 0xFF. Plan for:

  • requests from several TCP clients being queued and sent one at a time on the serial side;
  • a gateway serial timeout longer than the slowest device’s response, and a client timeout longer still;
  • a table of IP address, Unit Identifier and serial address for every device.

Which one should you choose for your panel?

Situation Usual choice Why
Energy meters or VFDs with only an RS-485 port RTU That is what the device offers; keep the trunk short and terminated
PLC with an Ethernet port and Modbus TCP server TCP No serial wiring; several clients can read it
Many RS-485 devices plus SCADA on Ethernet RTU on the bus, TCP above a gateway Keeps one master per bus
Remote site, few devices, cellular link RTU locally, data sent onward by an edge device Avoids exposing Modbus over the public network

For moving whole machines onto a plant system, OPC UA is often the better upper layer; see our guide to retrofitting legacy machines with OPC UA or Modbus.

Where does Embedos fit?

The Embedos 4G Ethernet Modbus gateway page lists Modbus RTU and Modbus TCP with RS-485, Ethernet, Wi-Fi and 4G interfaces.

Download the Modbus RTU vs TCP Guide

Get the fill-in selection and RS-485 wiring checklist by email. We send the link to your work email.


Decode your registers with the free Modbus register interpreter, and size your bus with the RS-485 termination and bias calculator.

Next step

Polling more than one device? Embedos Edge reads Modbus and pushes it to the cloud. Book a 20 min commissioning call.

Building a new plant? Write these choices into your tenders with our greenfield plant digital readiness checklist.

How to Build an OEE Baseline in Indian Plants Before You Buy an MES

 

To build an OEE baseline, pick one constraint machine or line, fix the definitions of planned time, ideal cycle time and a good part, then log downtime, output and rejects every shift, by hand if you have to.

Calculate OEE as availability × performance × quality and keep the loss breakdown, because that is what tells you whether MES software will fix anything. Many Indian plants can start with the machines, PLC counters and shift registers they already have.

Who this is for:

Plant heads, process excellence heads and strategy or transformation heads deciding whether and when to buy an MES, and the engineers who will collect the data.

Embedos blogs photo of a large hydraulic forging press with yellow handling equipment on a factory shop floor

What you'll be able to do after reading:

Set up a defensible OEE baseline on one line, read its loss breakdown, and turn it into an MES requirement that vendors must prove against.

What is an OEE baseline, and why do you need one before an MES purchase?

OEE is a measure; an MES is software. NIST IR 7911 describes MES activities as dispatching and executing production orders, collecting production data and analysing production performance. OEE is one of the numbers that analysis produces. In the ISA-95 model, MES sits at Level 3, manufacturing operations management. An OEE baseline is that number, with its three factors and its losses, measured the same way over a fixed period before anything changes. For a general definition, Wikipedia’s article on overall equipment effectiveness is useful further reading.

Without a baseline, an MES project has no yardstick. With one, you can ask a vendor which of your losses their system recovers and check the answer later against your own locked figure.

We don’t trust our current OEE numbers. Where do we start?

Start with definitions, not software. Most distrusted OEE numbers are calculated correctly from inputs that mean different things on different shifts:

  • Planned time is not fixed. One shift deducts tea breaks and the weekly PM, another does not.
  • The ideal cycle time is today’s average. That hides every slow cycle and makes performance look close to 100%.
  • Short stops are never written down. A two-minute jam that happens thirty times a shift is invisible in a log that only records breakdowns.
  • Rework is counted as good. The ISO 22400-2 quality rate counts only parts that were right the first time.
  • Machines are averaged. A simple average of five machines’ OEE says nothing about the line’s output.
What is the difference between OEE and MES?

OEE is a performance measure. An MES is software that runs and records production, and it can calculate OEE among many other things.

How do you calculate OEE for a machine?

NIST AMS 300-11 defines OEE as the product of availability, performance and quality. ISO 22400-2, the international standard for manufacturing KPIs, uses the same structure; the full ISO text is a paid document. A free paper by Kang and colleagues, hosted by NIST, explains the ISO 22400 time and quantity elements behind these formulas. We use the ISO 22400-2 quality rate, which counts only good parts and excludes rework. The NIST paper itself uses a “quality buy rate” that counts reworked parts as good, so check which definition a vendor uses.

Factor Formula Loss it shows
Availability Actual run time ÷ planned busy time Breakdowns, changeovers, waiting for material or operators
Performance (Ideal cycle time × total count) ÷ actual run time Slow cycles and short stops
Quality Good count ÷ total count Scrap and rework

A shortcut gives the same answer: good count × ideal cycle time ÷ planned busy time. If your two methods disagree, an input is wrong.

Worked example: one shift by hand

These numbers are illustrative, not a benchmark.

Embedos blogs diagram of OEE shift time losses for one illustrative shift: shift time 480 min minus 30 min planned downtime gives planned busy time 450 min; minus 60 min downtime gives run time 390 min; minus 65 min performance loss gives 325 min; minus 15 min quality loss gives 310 min fully productive time. Availability 390 ÷ 450 = 86.7%, performance (0.5 × 650) ÷ 390 = 83.3%, quality 620 ÷ 650 = 95.4%, OEE 68.9%.
  1. Shift length 480 minutes, with a 30 minute planned break. Planned busy time = 450 minutes.
  2. Logged unplanned downtime 60 minutes. Actual run time = 450 − 60 = 390 minutes.
  3. Ideal cycle time 30 seconds (0.5 minute) per part. Total count 650 parts. Good count 620 parts.
  4. Availability = 390 ÷ 450 = 86.7%.
  5. Performance = (0.5 × 650) ÷ 390 = 325 ÷ 390 = 83.3%.
  6. Quality = 620 ÷ 650 = 95.4%.
  7. OEE = 0.867 × 0.833 × 0.954 = 68.9%. Check: (620 × 0.5) ÷ 450 = 310 ÷ 450 = 68.9%.
Bucket Minutes How it is worked out
Availability loss 60 Unplanned downtime
Performance loss 65 390 run minutes − 325 minutes of ideal production
Quality loss 15 30 bad parts × 0.5 minute
Fully productive time 310 620 good parts × 0.5 minute
Total 450 Equals planned busy time

Here the biggest loss is speed, not breakdowns. A downtime-reason screen alone would not recover those 65 minutes, which is exactly the kind of finding that should shape an MES requirement.

Which definitions do you lock first: planned downtime, ideal cycle time and good parts?

  • Planned downtime. In the ISO 22400 elements explained in the NIST paper, planned operation time = planned busy time + planned downtime, and planned downtime includes scheduled breaks, meetings and maintenance. So planned downtime comes out before availability is calculated. Decide what counts as planned and write it down.
  • Ideal cycle time. Use the validated or nameplate cycle time per product, not the usual average. If performance comes out above 100%, the ideal cycle time has been set too long (too slow), or a count is wrong.
  • Good part. Right first time only. Count rework separately.
  • Downtime reasons. A short list, ideally under ten, plus “other”.

Get production, maintenance and quality to sign this one-page sheet before collecting data.

Does OEE include planned downtime?

In the ISO 22400 elements explained by NIST, planned downtime is removed before availability is calculated. Some plants include it; either works if written down and kept fixed.

Can OEE be more than 100%?

Not with correct inputs. Above 100% usually means the ideal cycle time is set too long, or a count is wrong.

Who should own the baseline?

Production or process engineering, with maintenance and quality signing the definitions. IT supports the data.

Which data fields do you collect, and where do they live on an Indian shop floor?

Field Where it often lives today Notes
Run and stop times Machine run signal, operator log Short stops need a signal; logs miss them
Total count PLC counter, machine counter, shift register PLC counters can often be read over Modbus
Good and reject count Quality register, operator log Record rework separately
Downtime reason Operator log, maintenance register Use the fixed reason list
Product or recipe running Production plan, PLC recipe number Needed to apply the right ideal cycle time

If a PLC counter is reachable, you can decode PLC counter registers with a free tool. Where a sensor signal stands in for a run state, for example a motor current transmitter, you may need to scale 4-20 mA signals correctly.

embedos-blogs matrix of OEE baseline data fields against where they often live today: run and stop times in the machine run signal and operator log; total count in the PLC counter, machine counter and shift register; good and reject count in the quality register and operator log; downtime reason in the operator log and maintenance register; product or recipe running in the production plan and PLC recipe number.

Can we baseline with manual logs, or do we need sensors first?

Method Effort to start Captures well Misses
Manual shift log Low Long stops, reasons, scrap Short stops, exact times
Manual log plus run/stop and count signals Moderate Exact run time, counts, short stops Reasons, unless operators tag them
Full MES High All of the above plus orders and quality steps Nothing in scope, if adopted

Start with logs; add signals where logs are weakest. Older machines may need a retrofit to give a run signal.

Which machines do we baseline first?

  • The constraint, the machine that sets the line’s output.
  • The machine your MES business case leans on most.
  • A machine where the shift team is willing to log honestly.
  • A machine where a run/stop signal is easy to reach.
Which data fields are essential to collect for OEE?

Planned busy time, unplanned downtime, ideal cycle time, total count and good count, plus the product running and downtime reasons.

Do we need to stop production to set up a baseline?

No. Manual logging needs no downtime. Adding a run/stop signal may need a short electrical job, which can go into an existing planned stop.

How long should you measure, and what do you exclude?

No standard fixes a baseline length, so a fixed number of days, 30 or any other, is a choice rather than a rule. A baseline means something once it covers your plant’s normal variation:

  • every shift and crew, including night shift;
  • at least one full cycle of your product mix and changeover types;
  • your usual maintenance pattern;
  • a weekly figure that has stopped moving much from week to week.

Exclude only what your definitions sheet excludes, such as declared planned downtime, and note one-off events (a power cut, a trial run) separately rather than deleting them. Historical data can serve as a baseline only if it was recorded with the same definitions; otherwise treat it as indicative and start a fresh log.

Can I use historical data as a baseline?

Only if it was recorded with the same definitions you will use later. Otherwise use it as a rough guide and start a clean log.

Embedos blogs photo of a 16,000 kN transfer press in a press shop hall, with an operator standing beside it

What is a good OEE number, and why not compare with “world class” figures?

ISO 22400 defines how to calculate OEE, not what it should be. Published “world class” figures come from different sources, industries and definitions, and none is set by a standard. Your baseline is only comparable with itself, measured the same way later.

What is a good OEE score in manufacturing?

There is no standard target. ISO 22400 defines the calculation, not a goal. Compare a machine against its own baseline.

How do you turn the baseline into an MES requirement and vendor scorecard?

Lock the baseline and give every vendor the same one-page summary: OEE, the three factors and minutes lost per bucket. Then score proposals like this:

Ask the vendor What good evidence looks like
Which of our loss buckets does your system address, and how? A specific mechanism per bucket, not a general percentage
Will you calculate OEE with our signed definitions? Yes, including the ISO 22400-2 quality rate without rework
How will you prove the improvement? Measured against our locked baseline, same definitions, same machines
What data do you need from our machines and ERP? A signal list and an integration plan, see MES vs ERP

How does a baseline compare in effort with a full MES?

A baseline needs a definitions sheet, a log, a spreadsheet and, at most, a few signals. An MES needs product and routing master data, an ERP integration, operator workflows, training and change management. The baseline fits inside one department, and it tells you whether the bigger project is aimed at your real losses. Planning a new plant instead? Put the same definitions into your greenfield plant digital readiness checklist.

Is the baseline wasted if we buy an MES later?

No. The definitions, downtime reasons and baseline figures become the MES configuration and the yardstick for judging it.

Where does Embedos fit?

For the baseline period, the hard part is often getting run and count signals off older machines. The Embedos Edge page describes the device as capturing machine and sensor data and sending it to cloud storage or a factory control room. Our pages on plant KPI dashboards for leadership and shift-level visibility for plant managers describe how we approach this.

Download the OEE Baseline Kit

Get the fill-in checklist by email. We send the link to your work email.


Next step

Want a clean OEE baseline on one line before you commit to an MES? We run a fixed-scope POC in one department. Email team@embedos.io to discuss an OEE baseline POC.

How do you keep proving data center noise compliance after the site is built?

To keep proving compliance, an operating data center needs continuous boundary logging that compares day and night Leq with the CPCB zone limits (CPCB Noise Rules 2000) while chillers, dry coolers and DG tests change the plant from hour to hour.

Log at fixed property line points facing the nearest receptors, tag every record with plant state, and keep the raw data ready for a complaint.

Who this is for: facility managers, EHS leads, community liaison staff and consultants supporting an operating data center in India.

Hub: where data center noise comes from.

Tell us what you need on site

Need this on your site? Tell us your requirement

One survey vs continuous compliance: what each is for

A snapshot answers one question at one time. It is the right tool for an acceptance test after commissioning, an annual check, or an attended visit after a complaint. It is usually done with a Class 1 meter, an operator on site, and a written report.

A continuous program answers a different question: does the site stay inside its limits on every night of the year, including the nights when two chillers are on and the DG is under test? It uses permanent or long-term loggers, covers both the day and night windows, and raises alarms before neighbours notice.

You need both. The prediction work belongs in predicting night noise before you build, and meter class, Leq, L90 and spectra are covered in Class 1 vs Class 2 sound level meters. This post covers what happens after handover.

Is a post-construction noise study enough for compliance?

It is a necessary first step, not the end. A post-construction study proves the site met its limits on the days it was measured, with the plant state of those days. Cooling load, weather, equipment wear and DG test schedules all change later, so ongoing logs are what show the result still holds.

What changes after the data center is live

The plant you predicted is not the plant you run.

  • Cooling stages with IT load and weather. More fans and compressors run on hot afternoons and as halls fill. See how cooling plant noise changes with load.
  • DG testing is intermittent and loud. A monthly or weekly test can dominate a boundary reading for its duration.
  • Failures shift load. When one fan fails, the others in the bank speed up, and the level and tone at the fence change.
  • Night ambient drops. Road and city background falls after 10:00 p.m., so the same plant is easier to hear at night.

Outside India, some local authorities now write post-construction studies, recurring tests or permanent boundary monitors into data center approvals. One example is the Fairfax County data center noise study requirements ( foreign context only, not an Indian rule). In India the compliance frame is the CPCB ambient schedule plus any condition in your own consent.

Why do data centers need continuous noise monitoring?

Because the noise changes all the time. Cooling plant follows load and weather, DG tests come and go, and night background is lower. One reading cannot show all of those states. Continuous logs show trends, catch step changes early and give you a record when someone asks.

Which numbers the continuous log must compare against

The Noise Pollution (Regulation and Control) Rules, 2000 set ambient limits in dB(A) Leq by zone (CPCB C1, Delhi Noise Limit

Zone Day Leq dB(A) Night Leq dB(A) Night is lower by
Industrial 75 70 5 dB
Commercial 65 55 10 dB
Residential 55 45 10 dB
Silence zone 50 40 10 dB

Day is 6:00 a.m. to 10:00 p.m. and night is 10:00 p.m. to 6:00 a.m. A silence zone covers not less than 100 m around hospitals, educational institutions and courts, as declared by the State Government. Rule 7(1) lets a person complain when the ambient standard is exceeded by 10 dB(A) or more, or when night restrictions are broken.

DG sets have a separate product limit. New sets up to 1000 kVA made on or after 1 January 2005 must not exceed 75 dB(A) at 1 m from the enclosure, and user-end enclosures and exhaust mufflers need at least 25 dB(A) insertion loss, or must meet the ambient standard, whichever is higher (CPCB DG noise limits C2 ; Delhi noise limit).

That limit sits at the set. Your boundary log is judged against the ambient table above. This is not legal advice; confirm the zone category and any consent condition with your consultant or SPCB.

What are the CPCB day and night ambient noise limits?

Industrial 75 day and 70 night, commercial 65 and 55, residential 55 and 45, silence zone 50 and 40, all in dB(A) Leq. Day runs 6:00 a.m. to 10:00 p.m. and night runs 10:00 p.m. to 6:00 a.m. Night is 5 dB lower in industrial areas and 10 dB lower in the other zones.

Embedos blogs embedos-blogs-continuous-boundary-log-day-night-cpcb-embedos: week of hourly boundary Leq against CPCB residential 55 day and 45 night lines

Post-construction check first, then the ongoing program

Start with one careful attended check once the site runs near its design load.

  1. Measure at the same boundary and receptor points used in the baseline, so results line up.
  2. Record plant state for every interval: chiller and dry cooler stages, DG on or off, IT load band.
  3. Cover both windows. A daytime pass alone tells you little about the 45 dB(A) residential night limit.
  4. Compare with the prediction from the impact assessment. Large gaps point to a source or path the model missed.
  5. Choose the permanent points for continuous logging from what you learned.

There is no Indian rule we can point to that makes a permanent boundary monitor mandatory for every data center. Some US ordinances ask for one; treat that as context, and check your own consent or SPCB conditions with your consultant. Many operators keep continuous logs anyway because they are the fastest way to answer questions.

How often should noise be checked after a data center starts up?

Do an attended check after startup at near design load, then keep continuous logs at the boundary. Add an attended Class 1 check after any major cooling or DG change, after a complaint, and on whatever schedule your consent or SPCB condition sets. US annual retest rules are context only.

Where to place permanent or long-term boundary monitors

  • Facing receptors. Put points on the property line facing the nearest dwellings, schools, hospitals or other silence zone receptors.
  • Out of the near field. A monitor next to a dry cooler measures the dry cooler, not the community exposure.
  • Height and weather. Follow the ambient protocol for microphone height and use an outdoor kit with windscreen and rain protection (CPCB Protocol for Ambient Level Noise Monitoring, MPCB host).
  • Avoid reflections and roads. Keep away from hard walls and road kerbs that would dominate the reading.
  • Power and access. A point you cannot service is a point that stops logging.

Embedos blogs embedos-blogs-permanent-boundary-noise-monitor-placement-embedos: plan view preferred permanent boundary monitors facing dwellings, avoid near-field plant

Where should permanent noise monitors sit at a data center?

On the property line facing the nearest receptors, away from the plant near field, at the height the ambient protocol sets, with outdoor weather protection and clear of reflecting walls and busy roads.

What a compliance-ready continuous log must capture

  • Time-stamped Leq for each hour, rolled up to the day and night windows.
  • L90 or a similar background descriptor, so you can separate plant from road.
  • dB(A) plus dB(C) or octave bands when tonal hum is the complaint. See why dBA can miss low frequency hum.
  • Meter make, serial number, class and calibration dates.
  • Weather notes: wind, rain, temperature.
  • Plant tags at the same timestamps: DG run, chiller or dry cooler stage, IT load band.
  • Raw exportable records kept for as long as your consent or legal advisor recommends.

The CPCB ambient protocol describes supervised sessions with 1 second sampling and day or night coverage for routine and complaint work. A permanent station extends that idea across weeks and months, and attended sessions still matter when a result is disputed.

What should a data center boundary noise log include?

Hourly Leq rolled into day and night windows, a background descriptor such as L90, dB(C) or bands for tonal issues, meter serial and calibration, weather, and plant state tags at the same timestamps. Keep the raw records.

Alarms that help operations before neighbours call

Useful alarms are simple:

  • Night margin. Rolling night Leq within a few dB of the zone limit.
  • Step change. A jump at one point that matches a fan bank going to high speed or a failed unit.
  • DG window. Flag test periods so readings are explained, and move tests into daytime where you can.
  • Data gap. A logger that stops is its own alarm.

Alarms are operational aids. They are not a legal finding, and the thresholds are yours to set. We do not quote fine amounts here; any penalty question belongs with your legal advisor and the SPCB.

Answering a neighbour or authority complaint with the log

  1. Acknowledge the complaint and note the date, time and location described.
  2. Pull the boundary traces for that window, plus the nearest façade point if you have one.
  3. Show plant state at the same timestamps.
  4. Compare the night or day Leq with the CPCB zone limit.
  5. Check tonal content if the words were hum, drone or throb.
  6. Fix source or path: fans and controls first (cooling plant), then barriers and enclosures when the log shows a path problem.
  7. Re-measure with an attended Class 1 check and share the before and after.

For why night levels matter to people living nearby, see health framing for neighbours. Night guidance from WHO Europe sets 40 dB Lnight outside as a target and 55 dB as an interim target (WHO Night Noise Guidelines 2009).

We make no health claim for any Embedos product.

How do you answer a data center noise complaint with data?

Pull the boundary log for the complaint window, show plant state, compare with the CPCB zone limit, check for tones, fix the source or path, then confirm with an attended Class 1 measurement.

Checklist: continuous compliance program for an operating Indian site

# Item Done
1 Zone category confirmed for each receptor (industrial, commercial, residential, silence) ☐
2 Post-startup attended check done at near design load, day and night ☐
3 Permanent boundary points chosen facing receptors, out of near field ☐
4 Meters: Class 1, or Class 2 stated as screening, with outdoor kit ☐
5 Day and night Leq plus L90 logged and retained ☐
6 Calibration schedule and field calibrator checks recorded ☐
7 Plant state tags (DG, cooling stages, IT load band) on the same timeline ☐
8 Alarm thresholds set for night margin, step change, data gap ☐
9 Complaint pack template ready ☐
10 Review after any major cooling or DG change ☐

Writing noise limits into purchases helps too: see writing noise limits into equipment tenders so the plant starts quieter. Inside the hall is a separate exposure problem: staff hearing in the data hall.

Can continuous logging replace a Class 1 attended survey?

No. Continuous logging is best for trends, alarms and answering “what happened that night”. An attended Class 1 survey, with a calibrated meter and an operator who can note what is audible, is what you bring to a dispute or a formal report.

Automation angle: third-party meter into Embedos Edge

A Class 1 or Class 2 sound level meter or noise recorder (IEC 61672, a paid standard) with an analogue output can be wired to an analogue input on Embedos Edge, using 4-20 mA or 0-10 V.

The Embedos Edge is the platform that logs that signal day and night, next to DG run and cooling stage status, and can raise an alarm when the night margin is at risk.

The Embedos Edge has no microphone of its own and does not measure sound by itself; the class belongs to the meter. Edge makes no compliance or health judgement.

To check how a meter’s range maps onto a 4-20 mA loop, try the 4-20 mA loop scaling calculator.

If a hosted view is needed, monitoring screens sign in at EmCommandView.

Embedos blogs embedos-blogs-slm-analogue-to-edge-embedos: third-party sound level meter analogue output 4-20 mA or 0-10 V into Embedos Edge platform, optional EmCommandView screens

Next step

Send us your site plan, receptor locations and meter model, and we will help you plan boundary logging points and the analogue wiring into Edge.

Email: team@embedos.io

Help Us Help You Leverage The Power Of Industrial Automation & Industrial IoT – IIoT.