A Thermal Imager

 Over the years, in fact decades, I could have made use of something other than my fingers to gauge the temperature of suspect elecronic components on circuit boards and, in watching YouTube videos it has reinforced my view that it's time I invested in a decent imager. These are not cheap but a few years back I reckon that what was available was either way too expensive or just not up to the job.

 

 Anyway, after searching for something suitable, I decided to buy a product that turned out to be not quite good enough. Let me explain... I decided to get a device which uses a mobile phone screen rather than its own display. The reasoning being that the cost of the display and its housing would surely detract from the quality of the infra-red sensor and its electronics. I picked the Mileseey TR256i which plugs into an Android phone. It uses a thermal sensor with a definition of 256 x 192 points and a scan rate of 25Hz. My cheapest option was AliExpress at around £136 including VAT with zero? duty. From Amazon the price is £199.99 (October 2024).

I duly paid a Chinese supplier and waited... and waited... and waited until giving up. The package had tracked into the UK but vanished after leaving Customs for Parcel Force. Something on the packaging had clearly attracted a thief so after the requisite 40 days I was reimbursed.

But in the meantime I'd investigated other products.

 

 

 I spotted the Infiray P2 Pro which offered a similar spec to the TR256i but was a lot more expensive. AliExpress had various suppliers averaging about £220 plus VAT (= £264) and Amazon circa £329. There was a drawback however... would it vanish en-route and/or would Customs add duty if purchased from China?

Also, to make the P2 Pro much more useful it's supplied with a macro lens reducing its focal length from a minimum of circa 400mm to around a few cm. This makes it perfect for close-up printed circuit board analysis.

There's a big problem. Pricewise it's not always easy to buy a device without the lens and if you manage to do this a lens may not be readily available. I decided to wait until Amazon had a special offer and a few days ago, sure enough, I found I could buy a P2 Pro for £239 (a saving of £90). At that price there was no lens but I'd noticed that I could buy one separately. The device duly arrived but, not only had the offer been removed, but lenses were no longer available from Amazon... very odd.

 
  

 

Here are a couple of pictures of the InfiRay (magnetic) lens taken from Ebay.

You'll notice they're completely different with the left example at around £30 and the right £60. Are both made by InfiRay or is one counterfeit?

Looking at adverts for the complete imager plus lens some have one example and some the other. I'm inclined to believe the right one is right and the left one is wrong.

I decided to but the more expensive option and this is due to arrive in a couple of weeks because there are zero UK-based suppliers.

 

 Having received the imager how did it perform? Bad news... although the software installed on my Motorola E13 phone the picture from the imager was absent... just a blank black area. I must admit though that I hadn't fully expected success because of various notes accompanying sales information.. "check for phone compatibility" was one note, and "works only with Android 10 or later". When I bought my phone I'd noticed it used the "Go" version of Android and it was pretty cheap. I now wonder if it was cheap because the slimmed down version of Android cost much less for Motorola than the full version? Clearly "Go" wouldn't support the P2 Pro even though it was Version 13.

 

My only realistic option was to buy a new Android phone and as luck would have it I spotted, on Gumtree from a postcode with the same BH23 as me, a Motorola G31. Not secondhand, but brand new in a sealed box for £60. I drove off to the seller's address, purchased the new phone, swapped my SIM and plugged in the P2 Pro, and it worked perfectly.

 

 

 Now.. a few tips. The key part of an imager is its thermal camera chip. Don't just pass a glance at the advertising material though because it's common for the display size rather than the imager thermal definition to be quoted to give one a false impression of the performance. A really good example is the Flir One Gen 3 pictured on the right.. "High resolution IR images with 1440 x 1080 visual resolution and 80 x 60 thermal resolution". The Hikmicro on the left (a handheld type using a built-in screen) looks good value but it's 240 x 240 spec is backed up by a 96 x 96 thermal chip.

 

Once you select the 256 x 192 infra-red chip prices can go sky high.

 

I suppose, with a little experimentation, one could find a suitable "macro" lens and carry out adequate circuit board analysis for an alternative choice of imager, but the P2 Pro magnetic lens seems to be an easy way to get this.

 

 

 This picture was my first test image. The video definition is 1440 x 1080 pixels with the 256 x 192 infra-red chip.

One of the reasons I bought an imager was to help identify a problem with our central heating and in this application I don't need a macro lens.

Our pipework is unfortunately buried in our concrete floor making it very difficult to maintain. For example, about 20 years ago, a damp patch appeared on our living room carpet around the time the central heating overflow pipe began to drip very hot water on visitors to our back door.

The overflow was caused by air entering a damaged copper pipe and displacing water which was being returned to the header tank. This was very hot water and eventually entered the overflow pipe in the side of the header tank which was routed to a point above our back door. In fact, previously if one attempted to bleed air from the system, the reverse happened and I'd corrected this by reversing the pump. I was then able to bleed the radiators but alas... water then escaped from the damaged copper pipe embedded in the concrete floor. Leaking water was emerging from a crack in the floor rather than directly above the pipe, but by using my metal detector I was able to pinpoint a right angle bend under the floor. This proved to be the source of the leak. It seems that when the pipes were installed back in 1983, the guy had bent the pipe too much and in lessening the angle acuteness had weakened the copper at the inside of the bend in the pipe. After many years of expansion and contraction metal fatigue had cracked the copper and resulted in the leak.

I repaired the leak by cutting out the bend and using a brass fitting.

 

All was then well... but a few years back various problems had developed. Currently, a section of the pipework seems to have got bunged up and the downstairs radiators usually run only between 20 to 30 degrees Centigrade. After much thought we installed a filter with a drain tap in place of our small kitchen radiator. This allows me to draw water from either the feed or return pipe.

By experimenting I discovered that drawing a couple of buckets of water from the return pipe drew hot water into the cool radiators. Essentially, this was because it enables the pump to better push water through the suspected bunged up pipes. I guess fitting a more powerful pump would also fix the problem, but a more powerful pump would need modification of pipework as pumps having a rating of more than 100 watts are physically larger and their width exceeds that in the existing pipework.

Anyway... the thermal imager should allow me to at least discover the location of the pipes.

The picture above shows floor heating from a pipe close to the boiler. This one is easily locatable... as you can see it's our cat's preferred location. But, is this a single pipe or both feed and return? Maybe the macro lens (when it arrives from China) will let me discover this?

Looks promising...

 

 

 Now to report progress .. the new macro lens finally arrived but, to my surprise, (looking at the pictures shown above) it appeared to be the fake rather than the genuine article. This was annoying to say the least because I'd deliberately paid full price for the genuine article. Without too much confidence I complained to the seller and, to my surprise, they said they'd refund the difference and, as the lens seems to work OK, I was happy.

Anyway I set to work to further investigate the central heating. There's good news and bad news. The good news is that i discovered a one-way valve near the boiler fitted the wrong way round. The bad news is that having fitted a new one the correct way round the heating shot up to the mid 20s Centigrade and our bills will now be sky high.

What else to report?

For many years my Teac AG15D Hi-Fi amplifier has had an intermittent problem. After getting very warm the audio would suddenly cut out so I'd fitted a large fan taken from an old PC. This is located in a gap just below the lid and worked fine in keeping the temperature below that where the sound would cut out. As the fan was relatively noisy I fitted a ballast resistor to limit its speed and this was fine. Now I have a thermal camera I decided to find out what was causing the amplifier to cut out. The first thing I noticed was the fan ballast resistor and then the device sitting on the video board. Although several parts were fairly warm nothing else was too hot and in fact those parts were common to the various output circuits. See the pictures below.

 

 

 

 My first observation was that fan ballast resistor was fighting the effect of the fan and heating up the space and secondly that hot device on the right is really too hot (bearing in mind the fan is running and the lid is off it could probably rise to well over 60C with the lid in place). Without the fan its temperature would probably rise to 100C so what is it exactly? I extracted the board and found it was a 7805F in a TO-220F plastic package. Looking at its spec with a passing glance (what someone else did in respect of a zener diode a long time ago at Plessey) this regulator will provide 5 volts at 1.5A, but that's not the whole story. If one studies the spec it reveals that max current is very dependent on temperature. The max junction temperature is stated as 125C and of course, because the junction is buried in the plastic case, the outer case will be cooler than this. Ordinarily semiconductors can run very hot, but in the case of this device, it's designed to shut down if the junction temperature reaches 125C which is clearly what was happening... So what can I do? The answer was simple.. replace the 7805F with a 7805AE which has a max junction temperature rating of 150C and has a much better response to the fitting of a heat sink. I found a small aluminium heatsink and fitted it to the 7805AE, removed the fan and refitted the lid. Now the Teac AG15D works properly and silently. It must remain a mystery as to why the amplifier first started to get too hot.
 
 

 Return to Reception