top of page

Lately I heard about the development of a condition monitoring project involving a team of more than 50. Most of the members were either in software and sales, and one, just one member on hardware. Given so many programmers I can only conclude that software is the new sexy.

The project is currently stalled because the choice of available hardware is either cheap and slow, fast and cost an arm and a leg, or buy basic unit cheap but pay for every little additional option which in the end still cost a lot.

This type of project requires a fast processing DAQ, which rules out models that do multiplexing and asynchronous modes of measurement. What is needed is a DAQ with high kHz processing speed for each channel. Such models are typically expensive leading to high prices which would not be welcomed by users especially in Asia where we love things cheep, cheep, cheap.

Fortunately, for this customer we have a solution for them. Something good, fast, and affordable. With our fast DAQ they can save a lot of processing cost which can come up to quite a pretty penny. This is good news for them because it will make their selling price viable to users.

The customer found it hard to believe that they never heard of our brand. Yet, our brand is not entirely unknown to them as they are already selling products manufactured using our brand of DAQ. Its just that their company is so big that its not just left hand does not know what right hand is doing, even the index finger does not know what the middle finger is doing.

If you are developing a project involving big data and machine learning in which the processing speed is critical but you cannot spend too much on hardware or be unviable talk to us. Send us an enquiry at

Vibration sensor reliability is important for all customers. I am sure everyone is familiar with the term GIGO which stands for Garbage In, Garbage Out.

Whether you are using the vibration sensor for data collection for the purpose of analysis or monitoring a machine 24/7 you would want a sensor that is reliable with a certain quality built that you can rely on and trust in.

Why would vibration sensor reliability be important?

1) Quality of data collected is important. More than ever today when the data collected is increasingly used to train AI-based software. You would want to have accurate data as it affects the predictive power of the algorithm you use to run the condition monitoring software.

If you are collecting the data for vibration analysis you would want data that will show as close as possible the true measured machine condition. In theory sensors of say 10% sensitivity from different brands should offer the same performance.

However, in practice this is not always the case as the quality of the components used in building the sensor can result in different levels of performance. Anyone can make a claim. You will only really know if this is true in practice when you are actually using the vibration sensor.

2) Your vibration monitoring of a machine can be disrupted by walkie talkie interference so it would pay to select a brand which is not prone to this. I recently attended a meeting in which the customer brought this issue up. This issue has been around for donkey years so I would think that by now it should be a non-issue but I guess certain things never get old.

3) Replacement of a vibration sensor can be costly especially if you are using a model with integral cable. I am not referring to the cost of the vibration sensor here but the cost of manpower to run the cabling not to mention the cost of the downtime and disruption to production schedule.

A bean counter will think that a $500 sensor is expensive (prices of sensors can vary from low to high depending on specifications) but when you consider that the average MTBF for a sensor is 25 years the cost of using the sensor would be $20 per year or $0.05 cents per day. And this is to protect an essential machine that in the event of a sudden breakdown can cost the company $XXX,000 per day (insert your figure here) in lost production.

If the machine were to be a person important to be protected would you pay peanuts to employ an old, out-of-shape, slower reacting bodyguard to protect the VIP? Or would you pay a premium to get a bodyguard that is fit for purpose? I am not against old bodyguards since I am no spring chicken either but it is what it is.

So if you would employ only the best for your VIP, why would you not do the same for your essential, must-die, die-not-never stop machine?

It is a basic fact of business that you get what you pay for and by extension, what you can live with. If you get it right the first time you save yourself some headache later on.

Contact us at if you are looking for quality but won't break the bank vibration sensors for your next project or simply to replace an existing sensor that did not live up to your expectation.

When you want to implement a monitoring solution how do you go about the purchase? Do you :-

  1. Get the vibration sensors, cables and controller from one vendor?

  2. Do you get each item separately from different vendors?

  3. Which item do you buy first? Do you verify compatibility before buying?

A recent encounter makes for an interesting case study in the buying process. A customer requested for a quote for 4-20 mA vibration transducers. In the same message the customer, Mr X, appended the specifications which is for a mV/g accelerometer.

I informed Mr X that if he wanted to use a 4-20 mA transducer he should be aware that it will not be compliant to the specifications put out by the plant. After explaining to Mr X, he agreed that he should use a mV/g accelerometer. I quoted him a transmitter that can convert the mV/g signal to 4-20 mA as per the specifications.

Then later Mr X said that he has the controller already. Before buying he wanted to test if the mV/g accelerometer can work with his controller. This was a bit unusual because normally the specifications of his controller should be able to answer this question. Or he could also check with the controller vendor.

Anyway, since his worksite is nearby I went to visit him and brought along an mV/g accelerometer. When I saw his controller I had a feeling that it should work with a 4-20 mA transducer but not with a 100 mV/g accelerometer. Anyway, Mr X connected the accelerometer and he could not get a reading.

On the spot I checked the controller's model on the internet and it was a temperature controller and no mention of it being able to connect to a voltage sensor, much less a vibration sensor. Mr X called the controller vendor and the vendor said the controller will work with a 4-20 mA transducer but he was unsure about the mV/g accelerometer as he was not familiar with it.

I asked Mr X if the controller he has is a sample unit that he bought to test or he has bought the entire quantity required. He replied that he has bought the quantity required. Mr X said he will ask the vendor to check if they have a controller that can work with a mV/g accelerometer.

And so the story ends here for the time being while Mr X is supposed to check with the controller vendor for an alternative model. Except a few days later Mr X has still not checked with the vendor. So I wrote to the vendor, sent him the specifications for a 4-20 mA transducer and a 100 mV/g accelerometer. And what do you know, the vendor quoted the same controller that Mr X has.

So it seems that the controller vendor was not paying attention as few days ago when Mr X spoke to him. Anyway, I replied back to the vendor and asked him to confirm that his controller can work with a mV/g accelerometer since the datasheet and manual that he sent did not specify this clearly.

This case has raised some points of interest :-


Why did Mr X bought a controller that was not in compliant with the specifications? In the aftermath of the testing, Mr X said that the output required is in 4-20 mA so perhaps that was the reason why he bought that controller.

I can agree with his thinking but in the event an audit is carried out and the controller and sensor is found to be non-compliant with the original tender specifications then Mr X runs the risk of rejection and having to replace all the controllers. I suggested to him that perhaps he should discuss with the enduser and get their agreement on using 4-20 mA transducer instead of mV/g accelerometers. Then Mr X's problem would be solved.


Why did Mr X not get 1 unit controller to test first instead of buying outright the entire quantity? As things stands now he stands the risk of having to buy another batch of controllers if the user insists on compliance to the specifications.


Did Mr X send the specifications of the tender to the controller vendor to check, verify and confirm that it could be used with mV/g accelerometers before buying?


Mr X asked to test the vibration sensor after buying the controller. Did he ask the controller vendor to do a test to verify compliance to specifications? If no, then why not?

This case study illustrates the importance of checking with all vendors if one were to buy the controller and sensor from different parties.

As it is, I have this feeling that the controllers that Mr X has bought is going to be proved to be non-workable as far as complying to the tender specifications.

The best solution for Mr X is to have a discussion with the enduser and persuade them to accept the use of a 4-20 mA transducer while acknowledging the non-compliance with the published tender specifications.

I know the price of the controller is low but I would think that the vendor should take some responsibility to ensure that Mr X is getting the correct item. It would not be very nice to sell him something that he cannot use. I hope I can help Mr X settle his problem.

White Structure


IMS Systems Pte Ltd

48 Toh Guan Road East,

#07-109 Enterprise Hub, Singapore 608586


+ 6 5 - 9 1 8 2 6 4 4 3

bottom of page