need recommendation: Flir ETS320 vs TE-Q1

[pclion]

Hello everyone,

i want to buy a Thermal Camera for PCB inspection, which can u recommend me to buy ?
- Flir ETS320
- TE-Q1

thanks in advance!

[Vipitis]

Welcome!

I suggest you look through the forum a bit more and check out some other cameras as well. Down below I give you some links to get started.
So the two you listed are vastly different price range. If you want to spend about 2.8k€ on the FLIR - you can get the Thermal Expert TE-V1 and build a desk stand yourself fore cheaper. You will need to add a phone or computer to read it out and display it on a screen as well. The resolution of the V1 is greater than the FLIR ETS320 and that will give you a much better image.

Check the gallery to see some example images from the cameras you thought about: https://www.eevblog.com/forum/thermal-imaging/thermal-imaging-gallery/

If you want to learn about all the flaws of the ETS320 and why it is not recommended here, read: https://www.eevblog.com/forum/thermal-imaging/flir-ets320-design-comment-and-improvements-by-fraser/

If you want to see where other discussions on PCB inspection setups went, read and comment here: https://www.eevblog.com/forum/thermal-imaging/thermal-camera-for-smps-pcb-checking-and-finding-shorts-on-down-to-0402/

more links here: https://www.eevblog.com/forum/thermal-imaging/how-can-i-turn-my-ets320-into-a-real-microscope/


Don't be shy to ask further questions.

[frenky]

I can only give opinion about T.E. Q1 with 13mm lens.

It has manual focus by rotating lens and you can focus really close (see photos).
You can use it connected to PC and use i3system windows software or JoeC windows software which has many more features.
Or just use it attached to phone which is very convenient (to me).

USB dongle with voltage regulator heating up:





If you try to focus any closer the lens will fall out because you are unscrewing lens with focusing closer and closer...

[Fraser]

Regarding getting even closer focus with the TE-Q1...

In industrial CCTV circles it is normal to use spacer rings and extension tubes to achieve close focus when using a standard lens. In the case of the Q1 I would investigate whether the lens thread is a standard photographic/scientific diameter and pitch. If so it MAY be possible to find a short extension tube that is basically a male to female adapter. Those with a lathe or access to an engineering facility could possibly make their own short extension tube.

Such a short extension tube could also be easily designed and created using a 3D printer 

Fraser

[mahony]

Regarding getting even closer focus with the TE-Q1...

In industrial CCTV circles it is normal to use spacer rings and extension tubes to achieve close focus when using a standard lens. In the case of the Q1 I would investigate whether the lens thread is a standard photographic/scientific diameter and pitch. If so it MAY be possible to find a short extension tube that is basically a male to female adapter. Those with a lathe or access to an engineering facility could possibly make their own short extension tube.

Such a short extension tube could also be easily designed and created using a 3D printer 

Fraser

Unfortunately it is not a standard thread - from what I measured it is M16x0.5 or a very similarly sized imperial format I am not aware of (any inputs to possible candidates are welcome here). But I assume it might actually be M16x0.5 because the also TE-V1 uses a metric format M25x0.5. As the thread is very fine I am afraid 3D printing might not be an option here but trying wont hurt either.

Probably of more use to the original question: From my opinion the TE-Q1 is a very nice camera for the given price tag, altough the software (still) really limits the image quality. On the other hand you get an SDK which you can use for whatever you want to and there is additional software (joe-c!). For the budget of an ETS320 you can even get the 640x480 TE-V1 - so 4x the resolution / FOV and if you like also with a fast frame rate - the Android App for it is even more anoying then the Q1's. In the gallery I have posted a couple of close up shots with the V1 but they do not use the original lens...

For PCB work only I would suggest the Q1 plus some good stand where you can adjust height and position of the camera and/or the PCB to get the best focus without always having to screw the lens in and out.  And probably also get a cheap dedicated Windows Tablet / Laptop to run joe-c software or the native one from i3. Still cheaper than the ETS and certainly not less capable.

Depending on availability the Opgal ThermApp might be another option - imaging quality is similar to the Q1 at a slightly higher price tag.

[Fraser]

@Mohony,

Thanks for the thread detail. As you say, a 0.5mm thread pitch is very fine even when 3D printing at 0.1mm layers. My ETS320 lens holders used a 1mm thread pitch which did print OK with 0.1mm layers. I could also print the threads at 0.2mm per layer and they still worked fine. Much depends upon the printer quality.

I had a quick look for photographic uses of the M16 0.5mm pitch threads. I did find a Chinese lens mount that uses such so it may be some odd PCB camera lens mount thread standard, like the common M12 x 1mm S mount.

https://www.aliexpress.com/item/5-100mm-Lens-3-55-Degree-MTV-CS-Mount-Infrared-Night-Vision-Manual-Zoom-Auto-IRIS/32792822447.html

If I found no other options, I would buy that lens plus mount assembly and extract sections of the male and female threaded sections to create a short male to female adapter.

Fraser

[frenky]

Not to confuse OP with this custom extenders for lens debate.
There is absolutely no need for that in pcb inspection. With min possible focus distance total width of subject in focus is 14mm. (A bit more than 1/2".)

So if you are not inspecting microscopical PCBs that is more than enough "zoom".

But for fun it would be interesting to see thermal macro images of electronics (inside IC) or nature using extending tubes for lens.
For making extending tube from plastics or aluminium this would be handy:
https://www.ebay.co.uk/itm/New-1pc-HSS-Machine-M16-X-0-5mm-Plug-Tap-and-1pc-M16-X-0-5mm-Die-Threading-Tool-/292393772416

[Fraser]

Franky,

Good point. When 'good enough' IS enough   

Some users are looking for a thermal 'microscope' but, as you state, most PCB inspection does not need such magnification, even with today's tiny SMT devices. I was just looking to address the 'not enough thread's issue when close focussing the standard lens on the Q1    The pictures that have already been provided show very nice close focus capability 

Fraser

[eKretz]

There is no inch equivalent to an M16 - 0.5 thread. The diameter is close to 5/8" but the pitch would equate to 50.8 threads per inch. 50 threads per inch would probably work for short thread engagements but 5/8" - 50 is definitely not an off the shelf thread. Many equipment makers over the years have made proprietary threads that aren't to any standard - in the U.S. system these would usually be categorized as UNS - Unified Special - as opposed to standard thread sizes which are known as UNC (Unified Coarse) or UNF (Unified Fine).

My guess is that the thread in question is just exactly what was supposed from the start - an M16 - 0.5.

[frenky]

I have just measured it. It's exactly M16 x 0.5mm.

[Vipitis]

while not being beautiful, you can print extension rings without threads and use use tension, duct tape or rubber bands to fix it. It won't allow for fine adjustments, but on a desk stamp made from something with nice height adjustment like a microscope, it will work. The DIY auxiliary lens I made allows me to extend the front element nearly twice the usual distance because I made a thinking error in the 2nd prototype, but therefore it acts as a makroscope. which gave me the idea of printing the whole focussing mechanism yourself using larger threads that 3D print well - always do ISO metric - I learned that by wasting lots of time and filament.
It's not practical because there are other issues but I can try and get some extreme closeups if you guys have ideas for subjects.

but back to topic: do not spent your money on the FLIR ETS. It is not recommended on here and needs a few generations to improve in design. A much cheaper Thermal Expert - Q1 paired with some teststand and a windows tablet can give you a comparable if not better experience and if you like to tinker a little you will make it work for you.

[Fraser]

So, to help the OP with his decision........

The ETS320 is a novel unit that has some design flaws that need addressing. It can be greatly improved by some modification work on the part of the owner. Does it work OK in its supplied state ? Yes, but it is somewhat limited in its capabilities. For simple PCB thermal profiling, with a fixed focus distance from the PCB of around 70mm, it does work. Adding some decent software to the equation, such as FLIR ResearchIR 4 Max provides very useful thermal profiling functionality.

The TE-Q1 is a modern thermal imaging Dongle camera that requires a host, be it a smart phone or a PC. The processing power of the host dictates the frame rate of the camera but not the image quality produced. The TE-Q1 is intended to appeal to a market segment that is not involved in professional radiometric thermography. That is to say, radiometric accuracy may not be as good as professional industrial thermal imaging systems. It is easy to check the measurement accuracy of such a camera however. In many cases the TE-Q1 accuracy will be adequate. Much depends upon the needs of the user. The lenses used on the TE-Q1 are interchangeable and of far higher performance than the single lens option fitted to the ETS-320.

To put it bluntly, if I were mass producing a PCB and wanted reliable accurate thermal profiling of the PCB whilst operating, I would likely not use either the ETS320 or TE-Q1. I would be using a very capable industrial thermal camera with close-focus lens like my FLIR A40 or Jenoptik cameras.

For small scale PCB thermal profiling or diagnostics, both the ETS320 and the TE-Q1 will get the job done. The TE-Q1 is more versatile and offers the user far more deployment options as it is effectively a compact thermal camera head that may be deployed on an umbilical cable for ease of deployment. One issue to consider with the TE-Q1 is the analysis software. The quality of such software is crucial to the user experience when operating the camera. The software needs to offer the required measurement capabilities and decent image processing. This is essential for a dongle type camera as the ‘clever stuff’ happens in the host that is running the cameras software application.

Just some food for thought 

As has been previously stated in this thread...... sometimes “good enough” is enough and there is no need to over-spend. That being said, in the commercial world, professional grade equipment may cost a lot of money, but it can also pay for itself in terms of performance and efficiency in its tasking.

There is no better test than actually using a camera so, where possible, get a demonstration of a cameras capabilities before parting with your hard earned cash.

Fraser

[lolli-us]

First post on EEV, hi all!

I'm looking into buying a thermal camera for my research lab (budget ~1000 euros) to measure temperature gradients at the surface of water solutions. My key requirements are:

–Area to be monitored 40 to 60mm wide; to image it with the best possible resolution I thus need to focus down to a field of view that small.
–The temperature range that I'll be working with is very restricted, more or less 32-37C, and I should be able to detect temperature differences in the same field of view of 0.5C and above (I'm not interested in absolute temperature, but only relative temperature)

After reading several threads here I tentatively selected the TE-Q1 combo pack offered here in Europe as a good candidate. However, I have a couple of questions left and would be grateful for any feedback from owners of that camera:

1. Frenky, your photos of PCB closeups are fantastic, but the official TE specs say that focusing is limited to 0.4m-infinity for the 13mm/f1 lens. How did you get so close to get a 14mm field of view? Did you have to remove a focus limiter from the lens?

2. Does anyone have a photo of a water surface showing mild gradients in temperature, like the ones in my application? I'm really not sure of what I can expect once I get the camera as water is rarely the subject of pictures on forums...

3. What is the exposure time of the camera? (water eddies will look more smeared the longer the integration time)

thank in advance, Lorenzo

[Vipitis]

The Thermal Expert Q1 ist probably a good option. The Plus model gives you better accuracy because of the lens.

To focus closer you just have to unscrew the lens and move it a little further away, and mount it wit some kind of extension tube(which can just be cardboard). Alternatively you can add a close up lens in between - but that is a different topic and if you look through more posts you can find details.

Measuring water is a task I haven't heard about, but I can tell you that taking images of a hot liquid, like coffee does look rather nice and even a small temperature Delta of 5-8°C gives you good data.
I attach an image of hot tea, water with a little lower temperature should still work, I don't believe it will be reflective.

Exposure time is minimal, like 12ms on some cameras, you have to dig through i3 documentation to find it. But you are limited by the 9hz framerate.

[lolli-us]

Nice photo of the eddies in the cup of tea! Looks encouraging to me...

Reading Frenky's post here above I was left with the impression that he didn't need to add an extension tube. I'll search the forum for "TE-Q1 + extension tube" for more info and perhaps write him privately for confirmation. Thanks for the pointer.

On the exposure time the reason I asked is because on some older thread it was suggested that the reason why the TE-Q1 produces low noise images is because of long integration times. Couln't find any official specs on the TE website...

cheers, L

[eKretz]

I just made a few vids of hot water in the sink real quick for you. Should give you an idea of what the cameras can do.

https://youtu.be/7buYCOQyOTA

https://youtu.be/2tnA9PWjAvo

https://youtu.be/6Vto-vfi10M

https://youtu.be/2p44BBmL8EE

https://youtu.be/y-mk4-MnFLI

[Vipitis]

okay, the best I could find is the specs of a sensor they sell(the only one that matches resolution and pitch of the Q1(M1?): http://www.i3-thermalexpert.com/products/qvga-uncooled-infrared-detector/

it says "Thermal time constant 15ms" in a table when you scroll down.

[frenky]

1. Frenky, your photos of PCB closeups are fantastic, but the official TE specs say that focusing is limited to 0.4m-infinity for the 13mm/f1 lens. How did you get so close to get a 14mm field of view? Did you have to remove a focus limiter from the lens?

The is no focus "limiter". You unscrew lens to get closer focus and when the lens fall off the thread you are at min focus distance. 

This is min focus. I didn't use any extension tube just unscrewed lens completely and then carefully screw it back until the lens thread grabbed the lens.

[lolli-us]

The is no focus "limiter". You unscrew lens to get closer focus and when the lens fall off the thread you are at min focus distance.

Great news! Thanks for the feedback

[lolli-us]

I just made a few vids of hot water in the sink real quick for you. Should give you an idea of what the cameras can do.

Those vids are really cool... mesmerising. Thanks for taking the time to shoot and post! Really helpful.

[lolli-us]

okay, the best I could find is the specs of a sensor they sell(the only one that matches resolution and pitch of the Q1(M1?): http://www.i3-thermalexpert.com/products/qvga-uncooled-infrared-detector/

it says "Thermal time constant 15ms" in a table when you scroll down.

Thanks for pointing out that page: finally some specs assuming, as you rightly suggest, that it's the same sensor used in the Q1.

A thermal time constant of 15ms (which I would think poses an upper limit to imaging bandwidth) fits with the specified max frame rate of 60Hz, as the latter suggests an exposure time near 17ms.

My only remaining doubt is whether when the sensor is used at <9Hz the software integrates over a longer time frame.