Carrying Out A Wireless Site Survey


There are two main parts to a site survey. Firstly, a check of local conditions, secondly, a check of performance of the T24 equipment in the local conditions. This technical document is designed to guide you through the process to achieve optimal conditions.

Local Conditions

The first thing to do is to look around. Problems that you are looking for are physical obstructions and local 2.4 GHz radio use.


This could be anything either in between or near to the transmitter(s) and the receiver. Materials such as concrete and steel absorb and reflect radio signals. Also remember that water is a very good absorber of radio which means that trees in leaf can be very good at blocking the signal. Also, find out what may change when the transmitters are actually going to be used. Is there going to be a lot of vehicle traffic and will that create a moving obstruction leading to data loss?

Local 2.4 GHz radio use

To do this use the Spectrum Analyser part of the T24 Toolkit (ver 2.5.0 and up). You should monitor the local radio conditions both where the receiver is going to be positioned and where the transmitters are going to be positioned. The Toolkit uses the radio signal received by the base station so you can change the orientation to see if that has any effect.

Planar View Parts

Real-time Detected Signal The white trace shows the real-time level of detected signal. On its own this information only really indicates where other radios are operating. T24 works fine with other transmissions but you may want to stay away from channels that have a lot of activity when there are other quiet channels available.
Peak Detected Signal The shaded background shows the peak signal detected across the band. This is more useful than the real-time trace because, over time, this build a picture of where the traffic has the highest power.
Minimum Detected Signal The red trace is very important and shows the minimum signal level detected across the band. In a good, quiet RF environment these red traces will not be visible but where there is a high level of broadband noise or very high amounts of radio traffic you may see channels that show red areas. As long as these remain below the CCA (Clear Channel Assessment) thresholds for the T24 radio modules deployed (<=v3.x or >=v4.0) the T24 radios will still operate but given the choice select a channel that does not show a high minimum signal level. As levels start to increase above -95 db this will reduce maximum achievable radio range.
Band Noise Floor This indicates the lowest signal level across the entire band. Usually this will be off the bottom of the chart but when this is visible it can indicate underlying issues with the environment that could affect the T24 radio operation. As levels start to increase above -95db this will start to reduce maximum achievable radio range.
Radio v3.x CCA Threshold This orange dotted line indicates the signal level at which the version 3.x (and below) radio firmware will not transmit. Any signals detected larger than this level will stop the module from transmitting. Usually this is not a problem as T24 radio works in harmony with other radio systems and will transmit in the gaps between other radio transmissions. However, if the Minimum Detected Signal is close to, or above, this level then the T24 radio system will cease to function.
Radio v4.0 CCA Threshold Version 4.0 radio modules have a revised CCA threshold to allow them to work better in high noise RF environments.

Examples of RF Environments

This shows a good RF environment. The Band Noise Floor is low and there are no red traces indicating that there are plenty of signal free gaps to enable T24 to transmit. There is traffic across the whole band with higher signal traffic between channels 11 to 15, but there is nothing that would affect T24 operation.

Here we can see some visible red traces indicating the minimum signal levels. Around channel 2 there is something transmitting constantly but the signal is so low that T24 would operate fine anyway. However, channel 12 shows that there is a constant transmission that is above the v3.x radio CCA threshold so those T24 radios would not function on channel 12. Version 4.0 and above T24 radios would function but communications may be erratic and certainly the range and coverage would be reduced. It would not be a good idea to use channel 12.

Here we can see a scenario where the entire band noise floor is high. This means that across all channels the range achievable will be reduced because T24 transmissions from distant modules will be swamped by the constant signal from the noise floor. For most channels the minimum signal level is below the CCA threshold, so as long as the T24 signal is strong enough the system will still work. However, note the sloping nature of the red trace. At around channel 16 the minimum signal level is at the level of the v3.x radio CCA threshold so version 3.x radios would not be able to pair because channel 16 is used in the pairing negotiation. V4.0 radios would still operate successfully.

This shows how the display would look if the band noise floor slowly crept up. The red trace is only visible on channel 12 but other channels that were once OK (Having a very low minimum signal level) now have a viewable level of minimum signal noise. A double-click on the planar chart would reset the peak and minimum calculations so the minimum red trace would then follow the more recent higher noise floor – see image below.

Whilst you are doing this, it is a very good idea to take screen shots of the whole window (the spectral view can give a useful history as well). This will be a useful record and could help Mantracourt with fault finding. It will also be useful to note down what the noise floor was as this could limit the possible range.

Now it is a good idea to check how good the link is between the transmitter and receiver. There are two ways to do this. If you want to monitor multiple transmitters, go to the monitor page of the toolkit.

This allows you to monitor all transmitters on the selected channel. It is useful to check how many transmissions are getting through (with the Total column); you can compare the detected transmission rate with what you are expecting (big differences would suggest high radio traffic blocking sending the signal) The LQI column shows the link quality index which is essentially how good the radio signal is. This is in the range 0-100 where 100 is the best. Getting a reading of 80 will still give some headroom.

Double click the LQI heading to see RSSI/CV which gives some more detailed info.

Alternately you can pair to the module and look at the LQI and battery page. This will give you a historic graph of the LQI. Clicking the Advanced button also shows the RSSI and CV.

For more information contact us

Download PDF

Keeping our Fingers on the Pulse!

Recently we have been working hard to update the T24 pulse acquisition module. Our purpose with this update was to ease usability with sensors and improve upon the battery life when used as a counter.

Our initial brief was to look at the types of sensors that would generate a pulse output. The majority of these sensors seemed to have a 10 V minimum supply. They also tend to be open collector outputs (both npn or pnp). Another common use of the pulse module is to measure pulses from a reed switch and these switches don’t need the power supply to be anything more than the battery voltage.

One of the biggest improvements we have made is to improve the counter modes. The count mode is now interrupt driven and does not need the excitation to be on at all.

To complete our update we added some extra measurement modes. We have retained all of them from the previous module (frequency, count, rpm, interval) and added some extras:

  • Quadrature. We have added a second input line to the board and we now accept quadrature inputs.
  • Digital State. This mode has been designed to deliver the digital state of the input pin. The value is sent immediately on change and at the configured transmit interval. This is ideal for button press applications where occasional state change is sent to a controller.
  • Mark (Space). The input is now able to report the mark space ration of the incoming signal.

To meet these challenges in sensor types we have made all these settings configurable from within our T24 Toolkit without the need for any external components.

Configuration is simple, just select:

  • The type of measurement you need
  • Excitation – Choose between Off, 3, 5 or 12 Volts DC.
  • Pull Resistor – Choose to provide an on-board pull up or pull down (56 kΩ).

Then simply start using it.

For further details, advise and support please contact us at


This article was written by Dr. Clive Vallance, Senior Electronics Engineer, Mantracourt Electronics Ltd

Introduction to Digital Devices – Part 2


The previous instalment focused on the physical communication interface between digital devices and a receiving station. This time I would like to step up a layer and look at the information that is sent along the physical wires. This is often described as the communication protocol and is essentially the ‘language’ that is used to communicate.Print

I would like to maintain our analogy with the mobile phone interface a little longer. Recall last time we had two phones that were able to communicate and send audio information. This is akin to the physical wires in a digital conditioner. This audio connection does not care about whether the two individuals at each end of the phone understand each other, only that the connection is made. So, this extends a little further if we introduce a language that is being used to communicate. As long as the language is the same, the two parties are able to communicate and understand one another. This is exactly the same in the digital conditioner scenario.

Of course in our discussion we have not spoken about how people negotiate the use of the physical connection. If you have ever made a conference call you will understand how the complexity of getting onto that communication channel increases with the number of participants. If two people try and speak at the same time there is a collision, the information is lost or garbled and we instinctively cease communication and retry at a later point. In extreme circumstances someone may be allocated to control the access of the rest of the group to the communication channel which removes the chance of a collision. Alternatively, there may be one person requesting information from a number of participants in turn. All of these possibilities of negotiating access to a channel exist with a digital communication scenario.

In digital systems there is often one master who controls the communications between entities. This relationship is often referred to as master slave and is itself a mechanism of stopping collisions in the data. These systems often rely on the fact that there is only one master. Another arbitration scheme used in CAN relies on the priority level of a message. This is a little more involved but to be as simple as possible any message on a CAN BUS has a priority level. When a message is sent the message with the highest priority is allowed to transmit data by winning the arbitration phase. This is non-destructive and in a well designed system eliminates collisions. The CAN arbitration method does mean that a master is not required to control the access to the BUS. There are other access schemes which involve allocated time slots and tokens but I won’t dwell on those at this time.

At Mantracourt we maintain communication languages for different purposes and applications. Some are quick and others are easy to understand without a translating tool. Some are industry standardised and others are proprietary. In general we provide 5 main protocols.



ModBUS is a standardised protocol originally designed for communication between programmable logic controllers. It is versatile and robust and commonly used for industrial communication systems. It is simple to implement, openly published and royalty free. ModBUS RTU is probably the most commonly implemented variant of ModBUS and is used for serial communications. ModBUS TCP is now gaining momentum and is used across Ethernet networks although the real time performance is slightly more limited. ModBUS in general is quite efficient in its use of the communications channel as it is a binary protocol. Unfortunately it is not the easiest to read without a software tool to translate it. The protocol itself operates on a request response basis, so it is an example of our master slave implementation. We at Mantracourt offer digital devices that use ModBUS RTU on 232 or 485 serial interfaces.

I would also like to bring your attention to something we often come across in Modbus tools. Although it is standardised, the offset applied when parameter locations are read is a little ambiguous. I have come across numerous incarnations of PC and logic controller interfaces and they do have a tendency to vary the offset applied to a parameter. This is often the result of confusion between number starting from zero or starting from one. To be honest the pros and cons of zero and one addressing is a rant that I think I should keep for another day.

In conclusion, when the offsets are resolved the data across these networks is delivered reliably, robustly and efficiently. It is a protocol that I like to use. I think the greatest difficulty comes with readability which is often overcome with these translating tools.


This proprietary protocol is designed for simplicity of integration and debugging. The data can be connected through to a PC and displayed without any special software. A simple terminal program allows the user to evaluate the operation quickly and easily. This does come at the cost of overhead. The transmission of ASCII characters is not as efficient as the binary based protocols and as such it is limiting in terms of throughput. It is able to operate in two modes. Primarily the interface is request response, but there is also a streaming mode that exists. This then reduces the overhead in communication to increase the data rate but it does limit the operation to only a single sensor on the physical channel. As such it is not used often and has its limitations. Mantra-ASCII is available for use on 232 or 485 serial interfaces.

In conclusion the Mantra-ASCII communication is easy to debug and evaluate. As long as the data rate is not particularly high (circa 120 sample per second) it is a very simple and reliable protocol.


MantraBUS is one of our proprietary binary protocols designed to be efficient and flexible. It has much less overhead than the ASCII protocol and as such uses the physical layer bandwidth more efficiently. This leads to greater throughput on busy BUS interfaces. Of course as a manufacturer it is very useful because we have control over the interface and optimise it for our applications. This also means that it is much easier for us to support.

Unfortunately in most cases proprietary protocols do not play well with others (i.e. all the devices on the bus should communicate using the same language). As such system integrators have a choice. In the event that they are designing the entire system using the BUS then a proprietary protocol is fine and the protocol designer is able to offer lots of support. The flip side of this is that when the BUS is shared a standardised protocol is required and all the devices must use that same protocol.


CANopen is a data format that may be used on a CAN BUS. Often people are unsure of the difference between the CAN physical and protocol layers as the lines are a little blurred. CAN open is maintained by the CiA (CAN in Automation) and is one of a number of higher layer CAN protocols. It essentially adds network management controls on top of the CAN network. It is gaining wider usage within industry but does tend to be restricted to certain industry applications at the moment.

Our greatest struggle with CAN protocols is that each of the hardware vendors put together their own proprietary interface for a PC. As such configuration software, including our own, is often limited to specific hardware interfaces. There is another separation between CAN and the other BUS technologies that ought to be raised here. CAN is slightly separated from our analogy of phones and languages. We discussed briefly what the complexities are of gaining access to the BUS and how CAN systems achieve this. The fact that you have a CAN device will mean that it will cohabit on a CAN BUS and not cause any damage to the communications. However this does not mean that out of the box the system nodes will understand one another or know what they need to communicate and to whom. Effectively you will need to tell each node who it reports to and how that is done. This is where all the knowledge and expertise of the integrator is required as the efficiency of the entire system depends upon this configuration stage.


Mantra-CAN is our CAN based proprietary protocol designed to be flexible for end users. The key benefit to this is that it can be configured to deliver data onto CAN networks in a format chosen by the user. This is also a limitation as it can never be fully compliant with other protocols but is great at delivering data onto proprietary networks. Essentially it is a protocol designed to play nicely with others without really interacting fully. One thing to note is that the configuration of the device must take place before connecting to the BUS.

We have customers who swear by it and others who are nervous about designing a system that uses it. In general this all comes down to understanding what this kind of protocol can and cannot do. If you understand what it is designed to do, its great.

In summary, there are binary, ASCII and CAN protocols available. Each application may require one or other of these protocols. The first thing to understand is that if speed is important to you, a binary protocol is probably the way forward and if you just want to plug in and use a terminal interface then it is probably easier to use the ASCII type interface. I personally believe that the protocol is chosen for each scenario but I understand that many vendors like to choose one and specialise in it.



There are two aspects I think you should take from this if nothing else. When using digital conditioners, before you get to the software you are using, ensure the physical wires and the protocol are the same between all of the devices you are connecting together. This is where 80% of the problems arise in digital implementations.

Next time I would like to move higher up through the layers again and look at the user interface software. This can be completely customised, use our interface drivers or even our free software. It all depends upon the complexity of the application but I will try and take you through a few options.

For more detailed advice and support please contact us at



This article was written by Dr. Clive Vallance, Senior Electronics Engineer, Mantracourt Electronics Ltd


Download PDF



Introduction to Digital Devices


We are well aware that when customers begin to use digital conditioners that they start on a journey of learning a new language. There are many advantages to using digital solutions, but there is a lot to learn, and the prospect of spending time learning to use a new technology can rule out these products when timescales are a factor. This blog is the first entry of an introduction to the language used in digital conditioners.

Mantracourt provide a range of signal converter modules which take load cell or strain bridge inputs and provide a data output in a variety of bus and protocol formats.

Most of our customers are quite happy with the concept of analogue signals which are output from signal conditioners and the methods of using that signal to display or log information. With this in mind I am only discussing the output from the conditioner and the ways that this data is transferred to the user interface or display.

This entry will focus on the physical connection. This connection can be on a signal bus (i.e. multiple devices connected to the same set of wires) or a simple point to point link. The useful aspect of the physical connection is that as long as the physical connections use the same interface it doesn’t care about what the data is that is sent over the link.

As an analogy, it’s quite readily accepted that two mobile phones can connect to each other and send an audio signal. It doesn’t care about the language that you speak, just that the protocol between the handsets match. However if we try to use a handheld walkie talkie we won’t be able to communicate with the mobile phone. Essentially the two handsets are sending the same audio information but because the communication methods are different it is not possible to pass on any information. This is the same for the physical interface between digital devices. If you have an output that is of one type at the sensor, it must be the same at the receiver.

In our products, we focus on four physical interfaces:


Lots of our customers request controller area network (CAN) compatible products. Often people are unsure of the difference between the CAN serial interface and the protocol that sits on top of that interface. Physically CAN nodes communicate using a differential balanced pair of wires and it is designed to be robust and reliable in busy and noisy environments. A common observation with CAN is that users don’t realise that it requires a termination resistor at each end of the bus in order to function properly. It is heavily used in the automotive sector and is gaining traction in other sectors (i.e. control systems).

There are certain aspects of CAN that make it unique in our list of physical connections. Unfortunately the drivers required for use on a PC tend to be proprietary and this is probably the biggest drawback when supporting CAN on a PC. This means that when we provide software to configure the product it will only work with a specific hardware adapter (in our case ixxat). Of course it is still possible to use other hardware interfaces but it will require the user to develop their own custom software in order to configure and talk using that interface.


USB is commonly used on PCs and mobile platform devices. It is probably the easiest platform to use but of course has its limitations. It is often used in office, test and laboratory applications as it is so simple. The interface itself allows us to provide software that can automatically connect to a device, configure and collect data. It removes many of the more difficult aspects which arise when customers are introduced to digital conditioners as many of the configuration steps in the other options are hidden away.

The problem with USB is that it will only work on cable lengths of less than 5 metres and it is not generally used to connect to industrial controllers. There is also a limit on the data rate that can be achieved but this is often related to the PC performance rather than USB. Another note of caution is in the use of USB Hubs. For reasons I won’t dwell on they often cause trouble. Be careful when using HUBS and try to avoid daisy chaining them together.


RS485 is a serial interface suitable for signal busses. As with CAN this serial format also uses a differential balanced signal pair to communicate and it needs to have termination resistors at each end of the bus (even in single point to point applications). It is capable of very long links (1200 m at 100 kB/s) and is commonly used in industry for logic controllers. Unfortunately there is an added complication with RS-485 where it can be used in either a 2-wire (Half duplex) or 4-wire mode (Full Duplex). This often changes a configuration setting on the serial interface hardware. It is widely available and the connection to a PC is standardised. Most serial port interfaces simply show up as a COM port on your PC. This means that we can be much more flexible with our configuration software on this interface. As far as a PC is concerned it is presented exactly the same as an RS232 connection.

RS485 is ideal for industrial applications as it has a good degree of fault tolerance and noise immunity.


RS232 has been the standard point to point communication channel for what seems like forever. The interface to a PC is the same as RS485 although the hardware is different (either a different setting on the hardware or a different device, depending upon the vendor). It is very simple but it can only operate as a single point to point link. As such it is often used for evaluation and doesn’t require the termination resistors we have talked about.  RS232 is suitable for short cable lengths up to 15m and is not as noise immune as RS485 or CAN interfaces.

So, we now have our communication interfaces. Each of the interfaces above can have different protocols, or in keeping with our mobile phone analogy languages, transmitted along the physical wires. As with any protocol there are pro’s and con’s associated with them. Some are very simple to implement and debug but they tend to be slow. Others are very quick but debugging and explaining their operation tends to be more challenging. We offer a number of industry standardised and proprietary protocols. The different protocols will be discussed in our next blog on this topic.

For more detailed advice and support please contact us at


This article was written by Dr. Clive Vallance, Senior Electronics Engineer, Mantracourt Electronics Ltd


Download PDF



Torque about…Ron Joyce

Let’s “Torque about”…

Ron Joyce
Analogue electronics and PCB design.



What is the strangest or most challenging job you’ve ever done?

I had a summer holiday job when I was at university working on the ‘new’ M5 motorway construction at Cullompton (it was about 1970!).

One of my tasks was to fill in the minute cracks in the concrete parapet on one of the bridges using water and a stone. I soon found out that I could get a much smoother finish (and enhanced job satisfaction!) by rubbing the water in with my fingers. After several hours of this, my fingerprints had all but disappeared and my fingertips were so sensitive I could have cracked any safe on the planet!

I promised myself not to do that again!

What would be your dream job?

Combining electronics and music on a full time basis.

If you had your time again, would you have followed the same career path?

Yes, I probably would. Electronics is both my career and hobby.

What are your passions outside of the workplace?

Music and guitars and guitars and guitar amplifiers and guitar effects … oh, did I mention guitars?

I’ve played in bands since I was at school and enjoyed every minute (not so much the travelling and packing up at the end of the gig though!).

However, note to younger self – practice more and GET SOME EAR PLUGS! I said GET SOME… oh, never mind….

I’ve have a close association with Manson Guitars, one of the best independent guitar makers in the UK if not the planet. This has led to me working closely with Matt Bellamy from Muse to bring his wacky ideas for new guitar effects to life (he uses Manson guitars exclusively). The most notable one being designing the X-Y MIDI controller touch screen fitted to his guitars. This always draws a great response from the audience, as does the guitar with laser beams firing off in all directions – most satisfying!

Another passion is, or rather was, motorcycling. Having ridden on and off (‘off’, quite literally a few times!) since I was 16, I’ve decided that having dodged the Grim Reaper more times than anyone could reasonably hope to, I’ve had a ‘sensible’ attack and decided to retire…. what am I saying!? Those Honda ‘Grom’ 125s look great fun!

My other passion is church bell ringing – oh, mustn’t forget my lovely wife Kim, my family, children and grandchildren – bless ‘em all.

If you could invite four people to a dinner party, living or dead, who would they be and why?

George Harrison my favourite Beatle. Anyone who dabbled in guitar playing in the 60’s wanted to be ‘George’.

Dan Denny the designer of the Vox AC30 guitar amplifier (The Beatles, Brian May, The Edge etc. etc. the list goes on and on). Arguably the most influential amplifier ever, pre-dating Marshalls by several years.

Jackson Browne, an amazing songwriter and that voice! I saw him play support in a pokey club in Plymouth in the early 70s, it was probably his first visit to the UK. There were only about fifteen people there and thirteen of them didn’t know who he was and didn’t seem that interested – their loss!

The Edge from U2. We could talk for hours about guitar effects and I could ask him why he always wears a hat – bald is beautiful, embrace it mate!

If they made a movie about your life, who would play you?

George Clooney of course.

What was the last album you bought?

Can’t really remember, it was probably something really obscure from the ‘bargain bin’. I hate to admit it but I’m a You Tube freak. The lack of top end in the audio doesn’t worry me as I can’t hear it anyway (see above passion).

Describe yourself in five words?

Peace-loving, quiet, dependable, not very good at counting…

Quick Fire Round

Tea or coffee? Both, depends on the time of day

Cat or dog? No preference. I love animals, that’s why I don’t eat them but I wouldn’t like one as a pet

The Beatles or The Rolling Stones? Both. The Beatles for their polished songs and the ‘Stones for their rough edge

Football or rugby? Rugby of course

Nightclub or restaurant? Restaurant – have you seen me dance?

Load Cell Fault Finding – Part 2


‘S’ Beam Load Cell

We often get asked for fault finding tips and so I’ve put together a simple fault finding guide for a load cell system.

In the previous blog we discussed the load cell part of the system, in this post we discuss the signal conditioning part of the system and provide a summary….

Check signal conditioner excitation voltage

If we measure the excitation voltage without the load cell connected we can check that it is the correct value for our load cell. A lower excitation voltage will give a smaller full scale reading but should still work correctly. A higher excitation voltage could cause the load cell to overheat and alter its characteristics. (If we then measure the excitation with the load cell connected this could show up other potential issues).

Check there aren’t too many load cells connected

We should also ensure that there are not too many load cells connected. Check the total resistance of the load cells either by calculation or measuring with a multimeter. Four 350 Ω load cells in parallel will drop the resistance to about 85 Ω. Check the specification on your signal conditioner to see what the minimum load is.

Short between signal inputs to signal conditioner to simulate zero load


Mantracourt’s Signal Conditioner ‘ALA5’

The next check on the list is to short between the inputs on the signal conditioner. This will simulate a load of zero and should give a zero output (with gain and offset set to one and zero respectively).

Ensure that shunt cal is not active

Some signal conditioners have a shunt cal function. This shunts the output by a fixed value to check the integrity of the load cell, wiring and signal conditioner. If there is any deviation from the last time shunt cal was performed then further investigation is required. If shunt cal is left on then it will completely mess up any real readings. Please note that the shunt cal should never be applied without the load cell connected.

Check six wire or four wire load cell and signal conditioner

If you have a 6-wire signal conditioner with a 4-wire load cell it is important that you link the sense and excitation terminals at the signal conditioner. Otherwise there will be no sense input and therefore no reference point for the signal conditioner. (If you have a 6-wire load cell with a 4-wire signal conditioner it is worth linking the sense and excitation as this will reduce the overall wire resistance).

Check correct power supply to signal conditioner

Check the output from the signal conditioner is what the display needs as its input. Some signal conditioners can output several different protocols which need to be set up correctly.


‘Pancake’ Load Cell

Ensure polarity is as required

Check that the positive direction of the load cell corresponds with the positive reading required.

Check gain and offset correctly set

Ensure that gain and offset are set correctly. It may be necessary to try recalibrating at this point if necessary, see the relevant manual. Depending on your setup this could be an absolute nightmare so should be held as a last resort.

The output cabling checks are very similar to the load cell cabling but will be marginally less prone to noise.

Most of the topics that I have covered here are very complex and I have barely scraped the surface. I’ve only very briefly touched on the output and display sections of the system which will be discussed in future blogs.

In Summary:

Check correct output type from signal

Check the signal conditioner has the correct power supply. See the relevant manual.

Load cell

  1. Check correct wiring connections and colour coding
  2. Check load cell resistance between excitation +/-
  3. Check load cell resistance between signal +/-
  4. Check for mechanical restriction on load cell movement
  5. Check to see if load cell has been overloaded


  1. Keep distance between load cell and signal conditioner to a minimum
  2. Keep all sensor cable runs away from inductive loads
  3. Ensure that you are using twisted pair, screened cable
  4. Check cable is screened and only connected to ground at one point
  5. Test cable continuity

Signal conditioner

  1. Check signal conditioner excitation voltage
  2. Check there aren’t too many load cells connected
  3. Short between inputs to signal conditioner to simulate zero load
  4. Ensure that shunt cal is not active
  5. Check six wire or four wire load cell and signal conditioner
  6. Check correct power supply to signal conditioner
  7. Check correct output type from signal conditioner to display
  8. Ensure polarity is as required
  9. Check gain and offset correctly set


Hopefully this has given a helpful starting point for fault finding load cell systems.

For more detailed advice and support please contact us at

By Tom Lilly, Technical Support Operator at Mantracourt

Load Cell Fault Finding – Part 1

We often get asked for fault finding tips and so I’ve put together a simple fault finding guide for a load cell system.

The output from a load cell is tiny! Any small interference will show up and then be amplified by the signal conditioner. The longer the cable the greater the interference will be due to increased coupling.  It will be improved if screened, twisted pair cable is used as with a good quality signal conditioner this will allow a lot of noise to be removed.

A certain amount of physical noise is normal, especially in very sensitive load cells. This can be from all sorts of sources ranging from computer fans on the same desk as the load cell to fork lift trucks driving around on the floor above! The more sensitive your equipment, the greater the degree of separation that is required from the rest of the world!

In figure 1 below, we can see a diagram of a load cell system.


Figure 1.  A load cell system

For this guide I have decided to split the fault finding into the areas load cell, signal conditioner and cabling.

I’ll start with the load cell…

Check correct wiring connections and colour coding

The first thing that I’d check is the wiring connections between the load cell and the signal conditioner. The colour codes vary between manufacturer and it is best not to assume anything. Usually the colour coding is on the load cell’s calibration certificate. Ensure that the wiring is as described. Getting this wrong can give inverted or biased results.

Check load cell resistance between excitation +/- and signal +/-

Also on the calibration cert should be the resistance between excitation +/- and also between the signal +/-. Measure those resistances with a multimeter to confirm that there is no internal damage to the sensor. (There will be a little variation from the figures stated on the calibration and this is normal). Don’t forget to disconnect the load cell from the signal conditioner before doing this.

Check for mechanical restriction on load cell movement

Have a good look at the load cell. Is it free to move? Any mechanical restriction will have an effect on the load cell’s output. (You can also use this phenomenon to limit the total movement of the loadcell to prevent accidental overload). Even poor cable routing can twist a small load cell and give skewed readings. Therefore it is important to make sure that the cable routing is free from any strain. Also the weighing vessel needs to be free to move in the direction of the load.

Check to see if load cell has been overloadedosbst-200x200

Another possible problem is if the load cell has been overloaded. A small load cell could have been overloaded simply by leaning on it. If this has happened it may have permanently deformed the load cell and it will not return to its zero position correctly. This is likely to show as a non-zero reading under no load. There is a high chance that this will be terminal! Talk to your load cell manufacturer for more information.

Now we can look at the load cell cabling. As previously mentioned, the voltage outputs from the load cell are very small and easily influenced by the outside environment. There are several ways of minimising the disruption.

Keep distance between load cell and signal conditioner to a minimum

The further apart the load cell and signal conditioner are, the more susceptible the cabling is to capacitive coupling and electromagnetic fields. Keep this cable short and there will be less interference.

Keep all sensor cable runs away from inductive loads

It is very good practice to ensure that all sensor related cables including sensor power are kept separate from any inductive loads and their power feeds.

Ensure that you are using twisted pair, screened cable

Using twisted pair cable can help reduce the magnetic influence of power cables. This needs to be in partnership with a signal conditioner’s common mode rejection capability. Screened cable prevents capacitive coupling by coupling to ground instead of the signal wires.

Check cable is screened and only connected to ground at one point

If the screen is grounded at more than one point a ground loop is created. This is the cause of the earth hum in audio systems. This is just as problematic in instrumentation systems. Remove any extra link(s) between screen and earth. DO NOT disconnect the earth wire of the appliance in question. This is dangerous and can cause serious injury.

Test cable continuity

A simple continuity test can check the integrity of the cabling. Disconnect the cable and check the resistance.
In the next post we will discuss the signal conditioning part of the system….

By Tom Lilly, Technical Support Operator at Mantracourt

Continue to Load Cell Fault Finding – Part 2

A Marketing Intern’s Perspective

My name is Konstantin Delov and I’ve been part of the Mantracourt Marketing team for almost half a year now. I spent my summer working on improving our online presence and supporting our network of technical partners in the promotion of our products. I was fortunate enough to be around and participate in the preparation of two of our product launches – both T24 and BroadWeigh Updated Range.

In the next few lines, I’d like to reveal a unique side of Mantracourt that very few people can see, along with a short story on my participation at an international innovation competition.

Behind every great achievement stand a number of strong supporters. People’s attention is often focused on the final result, but that doesn’t diminish the importance of the bits that quietly, but consistently contribute to it. Just like our signal interface solutions that support a variety of industries, often remaining unnoticed.

A few months ago, I heard that same monotonous buzz indicating an incoming email. Though the content of the message was a bit more special than usual. I was being notified by Microsoft corporation that me and my Exeter University team had reached the World Wide finals of the biggest student competition there is – The Microsoft Imagine Cup. I was going to visit their HQ in Redmond, Seattle, play with HoloLens and pitch to some of the most influential figures in the corporation.


Sounds exciting, but it’s not the type of trip that you just go to. It tested our limits as a team and individuals, required us to burn the midnight oil on a frequent basis. It involved thorough planning and outstanding discipline. Selling an idea is not among the easiest of things.

Fortunately, the work paid off and my team became third worldwide in the innovation category!


And yes, it might have been me filling out the quotes for the press releases and smiling for the photos but, once again, without the skills that I had gained at Mantracourt things could have looked way different. The company taught me consistency. I managed to get a close look of how an organisation with worldwide distribution gets its message across and reaches its sales targets. You can’t simply learn these lessons in a lecture theatre.

In this day and age, Mantracourt is a rare example of a company that doesn’t just blindly follow numbers. It is a place where every employee, customer, supplier or contractor is respected as a valuable part of a team, rather than a production or sales robot. And this pays off. Every piece of electronics that comes out of our factory is manufactured and tested by people who want to see the company thrive. Every board that we sell is sought and needed by the customer. We always make sure that it will genuinely help them towards achieving their goal, be it for project or distribution purposes.

By Konstantin Delov, Marketing Intern at Mantracourt

Why Go Wireless?

There are many good reasons to switch to wireless for your sensor requirements. It almost seems an easier question to ask ‘why not?’ but I’ll start with ‘why?’

The Benefits…

The first, most obvious plus for wireless is the lack of wires!  With moving loads, trying to keep the cables neat, tidy and unbroken can be a pain. Depending on the installation, replacing damaged cables can be very tricky, time consuming and expensive. Fault finding on complex systems can also be very time consuming.

With a wireless system the initial installation process can be very quick and easy. Wireless sensors can be moved to different places if continuous monitoring is not required, saving costs.

All our transmitter modules can run on batteries, also removing the need for power cabling. Depending on sample time and frequency battery life of up to 5 years is achievable.

Devices can be easily swapped out if required. It is even possible to change the data tag of the new device to match the one being removed so none of the receivers need to be changed.

Extra devices can be easily added to the system. Only requiring simple reconfiguration of receiver devices.

Wireless Issues and How We Address Them

People often worry about the security of their data. When sending data wirelessly anyone with the correct receiver could also be picking up all the data. Fortunately we have a way around this with Group Keys. Once a Group Key is set, only devices using the same Group Key can ‘see’ the data from that device. (The Group Key offers millions of different combinations).

Another concern can be about interference from other devices. Firstly, due to the T24 error checking, if data is being displayed, it will be the correct data. There is a possibility that other radio traffic could interfere with T24 transmissions and we therefore provide within our Toolkit a spectrum analyser to check radio traffic on the 2.4GHz frequency band. If heavy traffic is discovered simply switch to a less congested channel.

Before installing any system it is important to test that it is correct for your uses. Our T24 Toolkit allows for monitoring of radio signal levels at both the device and the base station to allow full analysis of variation with movement and other environmental factors.

For our testing we use an open field site with the transmitter 3 m above ground and the receiver at 1.5 m above ground. Increasing either height will generally increase the range. Any obstacles within range will affect how the radio propagates and will change the range and coverage.

It is also worth noting that the positioning of the components that make up any wireless telemetry system always involves a degree of experimentation.  Time spent on this aspect will always ensure a successful outcome.

By Tom Lilly, Technical Support Operator at Mantracourt

Tom’s Tech Tips

The T24 data logging software and data mapping software is now available to download free of charge.  Tom Lilly, Technical Support Operator at Mantracourt, talks us through this intuitive application…

The full functionality of T24LOG100 is now available to anyone that wants it at no cost. It means that people can have a proper experiment with an important part of a system without any expenditure. This should allow people to discover how useful and easy the T24 system is. You don’t even need a base station to have a look through and change the sample projects. The tooltip help and main help files are really useful for getting to know the software.

T24LOG100_Montage-webAll the functionality is there.  We have up to 100 channels of display and data logging, trend charts, mapping functions allowing intuitive visual displays, report generation and web server function for viewing on multiple platforms all over the world!

What do we mean by 100 channels?  Well, these channels can either be T24 transmitter modules or mathematical functions based on the transmitters and/or other display channels. Each channel has the option of setting a peak or valley hold or varying filtration levels. All display channels can also have warning, underload and overload levels set.

All the display channels can be logged to a csv file. This can be triggered manually or automatically at intervals and opened in spreadsheet software such as MS Excel.

Double clicking on any of the display channels shows detail about that channel including a trend chart with up to 10,000 data points.

The mapping function gives the possibility of creating intuitive displays needing very little explanation. Image files such as JPEG, PDF and DXF can be imported and other objects such as status icons, level indicators and displays can be placed on the page. With some clever thought, items can T24LOG100_Web_Server-webbe made opaque at certain levels and clear at others allowing the images to change as the load state changes. Up to eight pages of display mapping can be used. As previously mentioned, the example project has some useful demonstrations.

Reports can be generated and saved or printed when required. These can be in various pre saved formats.

Finally, my personal favourite is the web server. With this we can see a simplified version of the main display page in a web browser. This then means that if we know the IP address and password we can view this info on any computer, tablet or smartphone connected to the same network as the host computer. It is also possible that we can dial in from anywhere in the world to see the displayed data.

As well as the demo projects with the data logging software we now also have the first in a series of demonstration videos on LOG100. This can be found on the Mantracourt YouTube channel – click here to view.

By Tom Lilly, Technical Support Operator at Mantracourt