Author Topic: Use of a Thermal Camera for PCBA thermal profiling and repair  (Read 2916 times)

0 Members and 1 Guest are viewing this topic.

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Dianyang Technology has just released details of a new product that they are produced. It is a thermal camera system designed to analyze the thermal profile of PCBA's. Details are here:

https://www.eevblog.com/forum/thermal-imaging/shenzhen-dianyang-release-pcba-thermal-analyzer/

I already own a FLIR ETS320 thermal camera for PCBA work and thought it might make a nice topic for discussion here.

I will repeat my thoughts on what a thermal camera for PCBA work needs to offer the user and then move onto its uses in electronics design and repair. I hope others find this thread interesting and informative  :-+

It looks like I will be able to test the new Dianeyang Technology camera and I will review it on this forum.

Fraser
« Last Edit: June 05, 2021, 08:43:41 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #1 on: June 05, 2021, 05:01:30 pm »
Where PCBA inspection is concerned, we are in the realm of true thermography rather than ‘pretty pictures’ so key requirements are as follows....... (these are just generic and not specific to the Dianyang product)

1. Field of view .... is it adequate for the use scenario yet narrow enough to provide good IFOV ?

2. Manual focus ...... it is important to be able to focus the lens system for best clarity in the image. Fixed focus systems are a compromise solution that can mean poorly defined edge detail.

3. Thermal camera or Thermal microscope ? Such is determined by the lens system used. A relatively wide angle lens produces a thermal camera for PCBA overview but provides less detail. A narrower field of view combined with decent microbolomter resolution produces a thermal microscope that can provide good imagery of individual modern SMT components. Both types of imaging system require a close focus capability for PCBA work. With a single lens it can be hard to decide which type of PCBA inspection role to target. A wider angle lens is useful for a quick overview of a PCBA for hot spots or regions of interest. Digital Zoom adds little to the situation unless sophisticated interpolation is employed. A PCBA inspection and analysis  system will ideally provide at least two close focus fields of view. One for general PCBA assessment and one for detailed analysis of a region of interest. One method to achieve this in an affordable manner is the option to add a supplementary lens in front of the cameras primary lens. It is an old and well proven technique that can work well. Such techniques are common in stereo microscopes where supplemental lenses are screwed onto a mount in front of the objectives.

4. Resolution .... as usual... the higher the resolution provided, the better. It is worth considering the IFOV however as in this use scenario the correct selection of lens FOV still provides the required IFOV on target. A higher resolution would permit the use of a wider field of view lens for the same IFOV. I would consider 160 x 120 pixels the minimum resolution for PCBA inspection and 320 x 240 a good balance of resolution vs cost. The slightly lower Resolution Seek Pro cores would still provide adequate resolution in this scenario.

5. NETD ..... it is easy to get overly focussed on NETD when comparing thermal imaging systems. I advise caution as the NETD specifications can be misleading as I have discussed on this forum previously. When applying thermography to a PCBA inspection role, you need to consider the likely Delta T that will be present in the scene. In my experience NETD is less important in this application as the Delta T across a PCB is often quite large and so a relatively non demanding scene for a thermal camera. If trying to detect low currents passing through thin PCB tracks, you will likely struggle no matter what NETD is claimed. In high current failure modes, most components and tacks can glow brightly against their surroundings when observed by even a basic thermal imaging camera.

6. Measurement accuracy ? ..... let us get something out of the way immediately..... if you are thermal profiling a PCBA and need very accurate temperature measurements, you need to consider whether a thermal imaging camera is the most appropriate tool for the task. Most thermal imaging cameras state a measurement tolerance of at least +/-2C or 2% (whichever is greater) so you immediately have a potential error in any measurements taken on the PCBA. Then there is the issue of Emissivity. The emissivity of components on a PCBA can vary depending upon the material and shiny surfaces such as solder joints are a nightmare to measure with IR techniques. Plastic IC encapsulations are relatively simple to measure however. A thermal imaging system will help the user identify regions of interest or components that require further attention. It is sensible to use direct contact type temperature sensors on regions of interest to obtain more accurate temperature data. A unit like the Fluke Hydra equipped with many fine wire thermocouples is an appropriate temperature monitoring tool for the task.

7. System ergonomics..... just like when using a conventional microscope system, ergonomics are important for user acceptance. For a PCBA inspection thermal camera the user needs to have good visibility of the produced images. I would suggest that a small 3.5” LCD display, as found on the FLIR ETS320 PCBA camera is little more than an aiming aid ! Yes it works but a user will often prefer a much larger display. With this in mind, I am in favour of Thermal PCBA inspection systems that display their imagery on an external monitor of the users choosing, be that 10” or 100” diagonal dimension ! Granted the relatively low resolution of a thermal camera is an issue but a 12” to 22” monitor is a sensible choice for most tasks. This matter links in with PC connectivity in many cases as whilst some thermal cameras provide direct video output via Composite, VGA or HDMI, a PC is often used to analyse the images for more detail and measurements. With this in mind I support the format that uses a well designed imaging head that connects to a PC of the users choice and displays the images on a quality LCD panel, whilst also providing useful image enhancement and measurement capabilities through analysis software. As the owner of a FLIR ETS320 I can say that, whilst useful in a stand alone mode of operation, it really needs to be connected to a PC to get the best from it. The quality and format of the cameras mounting system is also very important to a user. Some mounting systems mimic a standard optical microscope design and can only accommodate a PCBA of of relatively small dimensions. For larger PCBA inspections there needs to be a “long arm” option that permits the user to mount the camera head on a long reach arm or an articulated arm. I would recommend that a manufacturer of such a product consider the head to stand mounting design to ensure that the camera head may be attached to either optional long reach arms or 3rd party articulated arm systems as commonly found on professional microscopes and even modern LCD monitors. The manufacturer could either offer alternative mounting options or ensure that the camera head is provided with a universal mounting system that may be used, or adapted for fitting to a 3rd party support system from microscope manufacturers etc. A PCBA inspection camera that can only accommodate the likes of mobile phone PCBA’s is severely limiting its potential l market appeal.

I will end this post here and move onto a new one to cover anything else plus software :)

Fraser
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #2 on: June 05, 2021, 05:03:07 pm »
OK.... part 2 of my general comments on PCBA inspection thermal cameras......

So far we have considered the camera head and the needs of the user where ergonomics are concerned. Basically a manufacturer needs to produce a versatile imaging solution that may easily be adapted to various roles found in both hobby and industrial use scenarios. Users have always had the option to mount a conventional thermal camera on a tripod or other mount to view a PCBA. Such is often a compromise solution however. The lens type and focus distance can be far from ideal for PCBA inspections and thermal profiling, but with little choice, the user make the best use of what they have available. A dedicated PCBA inspection camera offers those with a need for such with a ‘one stop’ ergonomically and technically refined solution to their PCBA thermal imaging needs. A hobbyist may elect to continue use of a generic thermal camera for such tasks due to cost but a PCBA inspection camera is still a viable product line as the electronics repair and research industries need such imaging equipment. High end thermal imaging equipment for PCBA work has traditionally been extremely expensive due to its specialist nature. The availability of more affordable imaging cores has the potential to change that situation.

As previously stated, I see a thermal camera based PCBA inspection system as comprising the camera head and a host that is running sophisticated image analysis software. Whilst a stand alone camera remains an effective tool, the added power of a host running analysis software adds much to the system, including the option for larger displays etc. The host system needs to be generic in nature fir broadest appeal in the market so I would expect to see support for PC, MAC, Linux, Android and iOS systems. The user may then select the host that most suits their needs. As a bare minimum, PC support is required.

The image analysis software is an area where a product can shine or fail. This applies to both all-in-one cameras and those using an external host fir image analysis. In its most basic firm, a hosts software will provide the user with the controls usually found on an all-in-one camera solution. Examples are....

Temperature Centre
Temperature Span
Auto Centre and Span mode
Emissivity
Ambient temperature
Distance to target
Colour Palette (LUT) selection
Spot temperature measurement
Image save
E-Zoom
Manual FFC activation

These are just the basics and image analysis software needs to offer a lot more to the user... examples are....

Spot and multi spot temperature measurement

Region of Interest

Highest and lowest temperature highlighting and measurement.

ISOTEMP capability

Scene Temperature alarm threshold and highlighting

Straight line temperature measurement plot

Image histogram view

Image stacking

Image annotation

Noise reduction algorithms with user selectable levels of effect

Mapping of temperature across the scene as a ‘heat map’

Temperature monitoring waterfall display

Electronic zoom with image enhancement through interpolation

Long term temperature logging of selected spots in scene

Camera measurement calibration the by user using reference Black Body sources (an advanced feature)

Video Recording function of display as seen by the user (spot temps, histograms etc.)

Time lapse image collection

Image saving options that include the RAW data, ‘As displayed’ frame grab, Radiometric JPEG and RGB JPEG.

Image overlay option for the semi transparent overlay of a reference PCBA board component layout (CAD image)

Image verification option for comparison of a DUT PCBA with a known good reference thermal image. Differences highlighted for user investigation and ‘alert’ differential values set by the user.

Option for a MSX style edge overlay using a visible light image of the PCBA taken with a separate camera.

PASS/FAIL PCBA test beaded upon preset permissible temperatures measures at specific points on the PCB. Many measurement points may be employed in such a test and thermal issues quickly identified.

Direct contact temperature measurement compatibility option where a thermocouple may be used to pass accurate temperature data to the software for use in its reporting. This would require a USB based thermocouple or PT100 thermometer as the data source. Such a data input could help with measurement accuracy and offset calculations by comparing the reported direct temperature measurement of an area with that reported by the thermal camera.

Fraser
« Last Edit: June 06, 2021, 03:09:54 am by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #3 on: June 05, 2021, 05:05:04 pm »
Use of a thermal camera for PCBA analysis

Thermal imaging and associated thermography has many applications in both domestic and Industrial applications. Some thermal cameras are designed to be a 'Generic' tool that may be employed in many applications with good results. Some specialist applications deserve their own dedicated thermal imaging solution however.
Such an application is Printed Circuit Board assembly thermal profiling and analysis.

In the electronics production and repair industry it is not uncommon to see thermal imaging in use to gain knowledge of heat distribution and concentrations of thermal energy on a PCBA. Such can help identify components that are underrated, stressed or inadequately cooled. Some faults on a PCBA will cause a hot spot at the point of failure or on a component that is being stressed by the fault. In other cases, the thermal image will indicate that a part of a PCBA is active when it should not be, or inactive when it should be. In production environments thermal profiling of a PCBA may be used in the testing phase to identify anomalies that require further investigation. Research and development teams may use thermal imaging as a means to evaluate the thermal stress, if any, present in a PCBA design and associated cooling solution. Such thermal profiling can improve product reliability and thermal management to avoid situation where either premature failure occurs or the cooling system is operating at excessive levels for user acceptance.

Let us look at the use of a PCBA inspection camera in real world scenarios to see how such might help us in our work. It goes without saying that the camera must be adequate for the purpose so I will not go into lens choices etc.

Example ..... R&D of a DSO and associated Switch Mode Power Supply PCB's.......

The DSO has been designed by the team to occupy a single multi layer PCB but uses a separate power supply PCBA for reasons of safety and thermal management.

The main DSO PCBA is placed under a thermal imaging camera system for thermal profiling. The PCBA is operating in the open air and not inside the DSIO casing. The thermal profiling involves monitoring the PCBA for 'hot spots' and then profiling those hot areas of the design to determine whether the temperatures are within expected and tolerable design limits. Any that exceed the design or component maximum temperature thresholds may be probe to premature failure and the cause should be investigated. In some cases the solution may be as simple as adding a heat sink to the component to dissipate the thermal energy into the ambient air. In more complex cases it may be that a component choice is underrated for the task or its deployment in the particular circuit is causing its stress. Such issues may require changes in components or how they are configured in order to reduce thermal stress. Once the PCBA has completed the thermal profiling stage and any associated rectification work, it is placed in its final location within the DSO casing. The same process is applied to the SMPSU to ensure that it is working within its thermal limits.

A thermal camera imaging friendly (thermally transparent) casing is used to carry out dynamic testing of the complete DSO design. The R&D prototype accurately reflect the intended production design in terms of ventilation and forces air flow around the internal components. This is very important ! It is at this stage of testing that the R&D team search for new 'hot spots' and monitor those that were previously profiles on the open test bench. The DSO may have forced air cooling and ventilation slots in the casing but this does not guarantee that the PCBA's will not suffer poor cooling through complex air movement within the case design. Designers often wish to use the lowest air movement possible within the casing to reduce fan noise that can irritate the user. variable speed fan systems are often employed to maintain the most appropriate fan speed for the current situation within the product. It is well known that in some designs the PCBA is perfectly 'happy' when operating in free air on an open bench, yet suffers thermal stress when enclosed within a casing, even when a case mounted fan is employed. the reasons are many, from air vortices caused by case shape to air flow masking of components by wide ribbon cables or safety insulating materials.
When ever a design is transferred into a new, enclosed environment, the PCBA and SMPSU should be checked for inadequate ventilation or areas of 'stale' air that become over heated resulting in a case hot spot or component stress in the area.
The fact that the SMPSU is located in the same case as the main PCBA brings its own thermal issues. Instead of the PCBA having its own cool air around it, it now has a PCBA that generates its own heat in the same air space. How the fan pulls or pushes air around inside the casing can move heat from the SMPSU to the air around the main PCBA or vice versa. Such can lead to the originally ambient air temperature around a PCBA rising to much higher levels and so poorer cooling of components.

The thermal profiling of a PCBA and the intended deployment within a casing can lead to design changes either to the PCBA cooling or the casing design. Basically any enclosed area of a PCBA or complete case design that may potentially cause a temperature rise in the components within should be tested for its effects on said components. An example would be a screening can on the PCBA to combat RFI that prevents the RFI issue but then generates an enclosed space in which 'hot' components cause elevated internal temperatures that stress either themselves or components around them. A reasonable solution that creates a new problem !If you cut ventilation holes in the screening can, the RFI problem returns. In such a case the designer would likely elect to use a mesh screen in the screened can for ventilation but this has its own issues, namely fine mesh clogging due to air moment and suspended dust particles. The overheating problem returns and the customer is not happy. Who said R&D was easy eh ?  :)


What about the electronics repair industry and its use of thermal imaging techniques ?

There appear to be two 'camps' in the electronics repair industry.... those who like to use the thermal camera as part of their fault tracing processes and benefit form it, and those who consider the thermal camera too expensive to justify and an unnecessary distraction form their well honed diagnostic process that work for them. There are some YouTube  video bloggers who discuss both sides of the argument in their videos. NorthridgeFIX is a video blogger who uses a FLIR Exx series camera to assist him in his investigations. He uses the technology to good effect. See a video of his that includes use of the excellent, but expensive, FLIR E60 camera....



A component that gets too hot is clearly either faulty or in distress due to a failure elsewhere. The thermal scene also quickly identifies areas of the PCBA that are getting hot and this can help identify what is, and is not working. In the case of a short circuit on a power rail, the thermal camera can be a very effective tool for identifying failed MLCC capacitors, ICs and layer shorts. I personally find a thermal imaging camera very useful when repairing electronics.

My very first use for the technology for a repair was many years ago when my girlfriend paid Sony to repair her shortwave radio. It came back to her working but after a few days she told me that it was eating batteries. I suspected that something in the radio was not switching off when it was supposed to. I opened the radio and observed the PCBA with a thermal imaging camera, an Agema THV550. It was immediately apparent that a section of the PCBA was still taking power from the batteries as it was warm. The Audio Power amplifier was the source of the heat. I traced the power amplifier supply rail back to a power MOSFET and was surprised to see a deliberate wire bridge across it, bypassing its control of the power rail. Upon further investigation I found the MOSFET to be open circuit so the SONY tech had clearly bypassed it as part of the fault tracing process and forgot to replace the MOSFET before returning the radio to the owner. The radio had been subjected to reverse supply polarity so some other components had been changed. The MOSFET bypass was just an oversight. A new MOSFET and normal operation of the radio was restored. I got some brownie points from the girlfriend for that repair :)

Thermal imaging enables the repair tech to 'see the unseen' and such can be a real aid to diagnostics. You can see a self reseting fuse activating before bringing a test probe anywhere near a PCBA. That little 'hint' can save valuable time and lead teh repair tech to the area wher trouble has occured. Our eyes are very good at spotting both change and patterns. If something looks out of place, our eyes may catch it.

Fraser
« Last Edit: June 05, 2021, 05:14:48 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #4 on: June 05, 2021, 05:29:47 pm »
Another NorthridgeFIX video containing 'thermal camera action'  :-+

Note that the FLIR E60 is a general purpose thermal camera and is not able to focus close enough to provide nice clear images of components. A supplemtary close-up lens is needed and steady hand, or mount, to support the camera. This is why dedicated PCBA inspection thermal cameras exist.

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #5 on: June 05, 2021, 05:34:52 pm »
And another nice NorthridgeFIX video showing the E60 mounted on a supporting articulated arm for a steady image. In this video a MOSFET lights up hot when power is applied to the laptop......

« Last Edit: June 05, 2021, 09:44:14 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #6 on: June 05, 2021, 05:39:08 pm »
NorthridgeFIX working on a dead iPad..... "signs of life" test with tha FLIR E60 thermal camera........

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #7 on: June 05, 2021, 05:42:38 pm »
NorthridgeFIX working on an iPad Pro using the FLIR E60 camera......

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #8 on: June 05, 2021, 05:49:20 pm »
NorthridgeFIX finds a tiny shorted MLCC capacitor on a 2015 Macbook Pro using the FLIR E60 thermal camera.... great work and demonstration of the benefit of the thermal camera in repair work.  :-+

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #9 on: June 05, 2021, 05:54:21 pm »
Not a repair but observation of very poor thermal management in the 2020 Macbook Air using the FLIR E60 by NorthridgeFIX.

This is the sort of issue tht should be spotted in the design stages using a thermal camera ! Apple really do surprise me sometimes.

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #10 on: June 05, 2021, 05:59:37 pm »
iPhone 6 GPU fault showing on FLIR E60 thermal camera by NorthridgeFIX

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #11 on: June 05, 2021, 06:03:03 pm »
LOVE THIS  :-+

NorthridgeFIX.... Alex smells burning in his lab  :scared:  He cannot locate by smell so uses his FLIR E60 to look for a heat signature..... guess what was burning up.... hint it was not a piece of equipment he was working on !

Also includes using the thermal camera on an Ipad PRO 12.9 .......

« Last Edit: June 05, 2021, 09:45:39 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #12 on: June 05, 2021, 06:06:44 pm »
Hard disk fault diagnosis using a FLIR E60 thermal camera by NorthridgeFIX

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #13 on: June 05, 2021, 06:08:21 pm »
HP laptop repaired in 5 minutes thanks to the Thermal camera  :-+

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #14 on: June 05, 2021, 06:11:10 pm »
2018 MAC Mini fault tracing using the FLIR E60 thermal camera

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #15 on: June 05, 2021, 06:14:36 pm »
iPhone8 fault diagnosis using the FLIR E60 camera  :-+

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #16 on: June 05, 2021, 06:17:22 pm »
Macbook Air capacitor short found with FLIR E60 camera.......

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #17 on: June 05, 2021, 06:32:10 pm »
Alex at NorthridgeFIX claims that the fastest way to find a short on a PCBA is to use the thermal camera....... iPhone 6 repair to prove the point......

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #18 on: June 06, 2021, 02:27:01 pm »
Just a quick post to clarify something…..

Alex at NorthridgeFIX owns a FLIR E60 and uses it to good effect in his electronics repair work. It is not essential to buy such an expensive thermal camera however ! I own an E40 upgraded to E60+ spec and I have used it to inspect PCBA’s. It is not really designed for that role however and it shows during use. The Exx series are quite large and unwieldy to hold over a PCBA so a supporting stand of some sort is recommended to maintain decent positioning and focus. The manual focus of the Exx series is an advantage over fixed focus cameras like the Ex series but the minimum focus distance is still not optimal for PCBA detailed analysis. A close-up lens may be fitted to improve the Exx series performance in this area. There are some excellent 3D printable supplementary lens holder designs for both the Ex and Exx series of cameras. The cameras remain quite cumbersome in use on the bench however as you need to look down on the display from above or angle the PCB.

This is where the introduction of compact ‘dongle’ cameras has improved matters greatly. ‘Dongle cams’ , as I choose to call them, offer the user a very compact camera head that is attached to a host computer/mobile phone. The camera head does the thermal data collection and the hosts processor does the ‘heavy lifting’ to process and display the data for the user. Dongle cams are now common in the marketplace thanks to the reduction in microbolometer prices and ‘explosion’ in the affordable thermal imaging product market that we have witnessed over the past few years. There are now a range of dongle cams offering greatly varying specifications and prices. Many offer good resolution, manual focus and decent optics. Temperature measurement capabilities are often offered but with varying levels of accuracy. These compact thermal dongle cams have much to offer the tech who wishes to construct a PCBA thermal analysis bench. Sadly life is rarely so simple though. Whilst generic dongle cams are good value for money, they can sometimes be limited in terms of their performance and the quality of software running on the host system. Some EEVBlog forum members have addressed the software issues by creating their own software with greater capabilities than that of the OEM. Sadly it appears that some OEM’s spend the development budget on the hardware and leave little for the software side of things ! As Thermal Dongle cameras are effectively “Software Defined Cameras” the quality of the firmware and software has a direct bearing on the end user ‘experience’ and product performance.

For dedicated PCBA analysis it is a good idea for the camera Design team to develop a compact thermal imaging camera head that collects an adequate quality of data for the task and passes that data to a host running software that is designed for PCBA analysis and provides desirable features for such use. There is absolutely nothing to stop the usual ‘generic’ use features being incorporated into the software as well. Such would produce a very versatile thermal imaging system for the lab bench that can perform PCBA thermal profiling duties plus more general thermography tasks such as viewing the environment within a products casing to thermal anomalies. The solution would need to offer decent images that enable the user to identify areas of interest on a PCBA. For this role close focus capability is important unless a very high resolution microbolometer is being used with a relatively narrow FOV lens. The PCBA inspection camera falls into an area of imaging between ‘generic thermal cameras’ and ‘thermal microscopes’. A close focus capability of around 60mm is usually adequate when using a camera with a 45 degree FOV and QVGA resolution. Even closer focus is clearly an added advantage. A 160 x 120 pixel microbolometer equipped camera would either need to focus at 30mm or use a ~22 degree FOV lens at 60mm to achieve the same level of PCBA detail. You need to consider the balance of microbolometer resolution against lens FOV and acceptable focus distance.

A small number of companies have produced dedicated PCBA thermal profiling solutions at affordable prices. The FLIR ETS320 is an example of such a camera system but whether you consider it affordably priced may depend upon whether you are using it in a high profit business or more humble environs ! I own one that I bought at a very good price but I would not pay the new price for it. The ETS320 is actually a FLIR E8 camera chassis bolted into a PCBA inspection format casing. The provided stand is just a repurposed digital microscope stand. The serious flaw in the FLIR offering is the use of the E8 fixed focus lens that has just been set to focus at around 75mm. That means the user must always maintain the working distance between the DUT and camera lens at 75mm. Manual focus would have greatly improved the ETS320 …. A real :palm: :palm: from FLIR on this front. The ET320 uses the standard generic FLIR Tools software and not a software crafted for PCBA inspection and thermal profiling. Another :palm: from FLIR.

There are PCBA thermal analysis cameras coming out of Asia that look very promising. Some use a 160 x 120 pixel microbolometer and I consider this the minimum resolution for most PCBA work and the lens FOV needs to be carefully considered. In PCBA thermal analysis it is the IFOV and resultant pixel size ‘on target’ that really matters as modern SMT components can be very small indeed ! With each year we see new and improved thermal imaging solutions coming out of Asia and the market continues to grow at a pace. The prices of such offerings are often very competitive when compared to western manufacturers of similar products. This is good news for the end users as competition in the market place is healthy and can lead to further price reductions. I will be very interested to test the DYT PCBA thermal analysis camera as it appears to offer both good build quality and performance…..at reasonable cost. From what I can see in the pictures provided by the OEM, the unit is basically a ‘dongle cam’ that has been designed for close working PCBA analysis and mounted on a versatile desk stand, yet may be tripod mounted and has a focus range of just 20mm to infinity :-+. I have yet to test the software package with the camera, but it appears to offer many useful PCBA centric analysis features. This could be what many of us have wanted for PCBA thermal profiling, but could not find at a reasonable cost. We shall see when I test it. Feeling positive about the product though.

Fraser




« Last Edit: June 06, 2021, 03:36:24 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #19 on: June 06, 2021, 04:13:01 pm »
In case anyone is wondering about visible light image fusion and FLIR MSX use on a PCBA thermal analysis camera…..

When looking at a PCBA in the thermal domain it can be hard to orientate yourself to identify the components of interest. The better the resolution of image ‘on target’ the easier it is to orientate yourself. This is one reason why FLIR introduced MSX …. The visible light based edge overlay assisted the user in making sense of the coloured ‘blobs’ they were seeing on the lower resolution cameras and gave the impression of greater resolution (good for marketing). The E4 in 80 x 60 pixel mode certainly benefitted from the additional information overlay ! For PCBA thermal analysis, if the microbolometer resolution, FOV, IFOV and pixels on target values are chosen wisely, there is little need for a live visible light image fusion or MSX edge detection overlay. I say ‘live’ as with PCBA analysis it is not uncommon to compare live imagery with saved reference data and images. Such reference material can include thermal images, visible light images and CAD PCB layouts.

When working at close focus distances the use of a visible light camera to capture image information for any form of overlay on the thermal image becomes challenging ! Not only must the parallax error be addressed but the visible light camera must also have the ability to close focus as well. On thermal cameras many such visible light cameras are simple fixed focus affairs. In this age of autofocus mobile phone cameras the issue can be addressed…at a cost to the OEM. In the case of FLIR and their ETS320 they decided, correctly, that MSX was inappropriate for their PCBA thermal profiling camera and they just deleted it from the BoM ! The ETS320 does not have the fixed focus visible light camera fitted, nor is it needed. When working at a distance of only a few tens of millimetres it is very hard to address the issue of parallax error and image distortion if the two cameras are angled to intersect at a set point. Yes software can do it, but OEM’s have realised that the benefits do not outweigh the costs and added complications of close focus VL overlays. The alternative has already been detailed in my ‘ideal’ specification at the start of this thread. The thermal image scene is captured and brought up on the users display. The display is either split in half or has the ability to show more than one layer simultaneously stacked. With a side by side format the ‘live’ thermal image may be compared to a stored reference image of a ‘known good’ PCBA and spot temperatures compared between the two. Temperature deviation alarms can be associated with specific areas of interest if they are configured by the user. It may also be possible to display a visible light image taken with a standard digital camera or digital microscope for side by side comparison with the thermal image to provide context to the thermal scene. A CAD type PCBA layout could be used instead of a visible light image, if desired. With a layer based display system the same visible light or CAD PCBA layout could be overlaid on the thermal image after resizing and alignment to match. This would create a fusion image but care is needed to set the overlay transparency levels correctly to avoid confusing the user with too much information.

When people reverse engineer a PCB, it is not common to photograph both sides of the PCB and digitally combine them on a computer display to create a two layer PCB track layout fusion image.  Such techniques may be applied in PCBA thermography to show which components or parts of the PCBA are associated with areas of thermal concern or interest. None of this is Rocket Science as users of image manipulation software will be aware. Layering images and using transparency settings can create a very useful, if sometimes complex, image for the user to study. The thermal image just becomes another layer of information….but it should take precedence over other layers in Thermal analysis work to avoid obscuring important thermal data from view.

Fraser
« Last Edit: June 06, 2021, 04:33:56 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #20 on: June 06, 2021, 06:14:42 pm »
Thermal camera mounting options for PCBA thermal analysis activities........

So we have a thermal camera that is suitable for PCBA thermal analysis but how do we mount it for greatest flexibility or to meet specific needs ?

It is at this point that I should make comment on a current trend in PCBA inspection camera usage.....in a word .... phones
That is to say, the mobile phone repair fraternity have woken up to the fact that a thermal imaging camera can be a VERY useful tool to them in their repair activities on modern electronics with its high density PCBA's containing miniscule components. Even probing such electronics safely can be demanding so a 'no contact' initial assessment can be of great benefit to the repair tech. Why do I mention this ? Well it may help to explain why many PCBA analysis thermal cameras are produced with a relatively small footprint base and compact support stand. Mobile phone PCB's are relatively small so the user does not need or want a large camera stand filling up their work area.

Not everyone buys a thermal camera for PCBA analysis on mobile phones however. Some of us work on much larger 'patients' and you only have to look at a Sony Playstation 5 or X-Box motherboard to see that teh small mobile phone centric stands are going to be somewhat challenged to cope with such large PCB's. That is not to say that they cannot be adapted or reconfigured to cope with such though. It would be a distinct market advantage if the OEM of a PCBA thermal analysis camera considered the needs of a wider user group in industry. If the camera may be removed from its more compact desk stand and used with either adapters or another stand to cope with large or tall PCB's, the camera becomes a far more useful tool. Achieving this goal is not dificult or expensive. Thr OEM just needs to capture the need when designing the thermal cameras mounting system. Something as simpe as a standard 1/4" 20tpi threaded tripod hole in a well chosen location can immediately off the user the option of adapters or 3rd party stands/arms for use with the thermal camera. Whilst a simple tripod mount hole may not mate with all types of support structures, it does present an easy to use mount for adapter plates and bracketry to adapt the camera mount to the needs of the stand, arm or bracket. The fact that FLIR omitted a standard 1/4" tripod socket on their well known E4 apalled me and when i challenged taht decison, I was told that the design team decided the Ex range was a portable handhelpd camera so did not need a fixed tripod mount  :palm:  When I saw thateh FLIR ETS320 also lacked any form of mounting adaptability like a tripod mount, I was similarly disappointed but the ETS320 is more like a microscope so it was possibly more understandable and a microscope arm arbor can be fitted to the ETS320 tube mount. Users of microscopes and devices like the PCBA analysis thermal camera can have varying needs when it comes to teh mounting method. OEM's should capture this requirement at the early development stages of a product and support it.

So wher does taht leave us ? Well in most cases any supporting system may be adapted to any camera, it just takes thought and adpatiosn to achieve  :-+ I am not adverse to making an adpater plate, angled bracket or spacer block but OEM's can make life easier or harder, depending upon how integrated their camera is to the supplied support system. a common source of support stands, articulated arms and all manner of unusual support structures, inclusing ceiling mounts are the Microscope, Photography and I.T. market segments. The microscope stands are the most likely best fit for use with a PCBA analysis thermal camera as the deployment needs are very similar, if not the same. A steady mount to hold an optical device in position near to a DUT with minimum fuss and good stability. Photographic camer stands also provide a vast array of mounting options and adapters. There are very solid copy stands trough to articulated arms, flexible arms and lightweight copy stands for portable use. The I'T' world may nor seem like a very likely source of mounting hardware for a thermal camera, but consider modern LCD monitor mounts .... you will now see all manner of flexible mounting solutions designed for VESA mounts on monitors and these are easily modified fro other uses. Articuated arms fro microscopes are often very expensive buyt a lightly used articulated arm for a lcd monitor is likley to be far cheaper. The key is to find a mounting method that best suits teh application and determine whether it may be adapted to hold the PCBA analysis camera. Be warned that some articulated arm munts have an expected end pint load rating that releates not only to safe use but minimum load for correct balancing against internal spring structures. With too little weight on the end of the arm, it may not stay where you place it and start climbing toward steh ceiling !

I attach some sample images grabbed from teh internet showing the vast array of microscope and bcanmera mounts that may be useful in a thermal imaging PCBA inspection bench application. prices vary wildly so shop smart !

Fraser

Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #21 on: June 06, 2021, 06:17:27 pm »
mounts continued....
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #22 on: June 06, 2021, 06:18:42 pm »
Mounts continued...
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #23 on: June 06, 2021, 06:22:32 pm »
So now that we have considered the matter of mounting a thermal camera above a PCBA, let us look at a few PCBA inspection cameras that exist in the market place and make you own mind up as to their adapabilty for alternative mounting methods......

Any that use a relatively standard tube mount as found on microscopes may often be adapted to mount on the drop tube (Arbor) of a microscope mount. They would need a tubular adpater to mount on camera or monitor mounts however. Those cameras that use flat bracketry to support the camera may be adaptable to take a standard tripod mount or may be fitted with a tube mount from a microsccope to gain microscope stand compatability.
« Last Edit: June 06, 2021, 06:30:12 pm by Fraser »
Cogito, ergo sum
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 11168
  • Country: gb
Re: Use of a Thermal Camera for PCBA thermal profiling and repair
« Reply #24 on: June 06, 2021, 06:27:56 pm »
Then there is the clever mind and abilities of the tech who adapts and creates what is needed........

(3D printing can be a great help in making adpaters etc)
Cogito, ergo sum
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf