Comparing affordable TICs

[torch]

At work, I am provided with a nice name brand unit -- 384 x 288, 3.5" display, 50hz, 35mK spectral resolution, -40 to 1,000°C measurement, intrinsically safe, NFPA 1801 certification, 2 hours looping video recording, etc. etc. etc. At darn near $20k, I can't afford that level of performance at home.

My home budget is more in the $500 buck range for occasional use -- automotive diagnosis, heat loss investigation, and, gee, maybe even circuit boards. I'm not interested in a smart phone dongle -- for one thing, I'm a dinosaur with a flip phone. So: a standalone unit. Stunned by the plethora of Chinese options and the untrustworthiness of their reviews, I attacked the problem on the basis of whether or not they were discussed favourably in this forum.

Actually, I rather hoped to find a chart similar to the ones on meters and scopes..

After a lot of reading, I think I've narrowed it down to 3 very similar units:

1. Hikmicro B20
Pros: Largest display at 3.2"
Longest focal length/narrowest FOV (disadvantage in certain applications)
WiFi

Cons: Memory is built in, not replaceable card
Wifi appears to only work with their smartphone app
Shortest battery life
No video recording
Most expensive -- about $250 more before taxes.

2. Guide PC210
Pros: Cheapest even after adding case and macro lens and shipping (but only by $1)
Lowest operating temperature, -15°C.
Only one with reflection adjustment
Longest battery life
Lightest (but only 5 grams lighter than the B20)


Cons: Smallest display screen, at 2.4"
Worst spectral resolution (but only slightly, 45mK vs 40mK of the other two)
Widest field of view (56° x 48°) (benefit in some instances)
Video recording possible, but only when tethered to a PC.

Topdon TC005.

Pros: Amazon can get one here in 2 days, free returns if it's defective.
Can record video internally

Cons: Lowest frequency (20Hz)
Narrowest spectral range (slightly: 8 - 14 um fs 7.5 - 14 um)
Only 2 span ranges available.
No adjustments for emissivity, reflection or distance
Heaviest (520g)

Unknown: IP spec, minimum focal distance, "fusion" or PIP modes?

Have I missed anything, or gotten anything wrong? Any other contenders I should be looking at?

[nikitasius]

> I'm a dinosaur with a flip phone.

well you can consider buying 50 bucks android phone w/ type-c TIC (and a handle) as a TIC w/ a huge screen. Why not? Honestly, i've checked the market and if i had flip phone it's cheaper to buy cheap android + type-c TIC (w/ a handle) to build a standalone device than to buy all-in-one TIC.

Thermal image will be pretty good and cover your needs, while temp measurements will be +- good. Also if you do diag you probably have other thermometers for focused & precision measurements and also it's even not the case to measure the temperature (you check for R/V/I, use your scope to check pulsations or whatever, even unsolder from pcb to check on tester or replace)..

Just my 2 cents

[nikitasius]

Also some topics from Fraser https://www.eevblog.com/forum/thermal-imaging/new-products-from-seek/msg5027698/#msg5027698

And you can google Infiray products and check real videos & teardowns on youtube.

[Fraser]

For “all in one” affordable complete thermal cameras you have never had more choice, but I can see how that could be a problem in itself !

I would suggest looking at the following models… not all will meet your needs but that will be clear from the specifications sheet.

In no particular order:

1. UNi-T UTi260B (UTi690B) - has a following on this forum but I consider it’s build quality marginal. Uses the Infiray Tiny1 C core.
2. Infiray C200 - uses the Infiray Tiny1C core and is a nice camera with good build quality.
3. Guide IR PC210 - uses their own Guide Sensmart TIMO256 core and is a well built camera
4. HTi HT18+ - Unknown which core is used but a decent performer from what I have heard. Not tested one myself.
5. HTi HT19 The next model up from HT18+ with higher resolution.

That will do for now. I personally would avoid going below 256 x 192 pixel resolution as that appears to be the entry point these days for a budget thermal camera with decent imaging capabilities. Many different brands of camera use the same common cores. Such cores are the Infiray Tiny1 and Guide Sensmart TIMO256. Hikvision appear to use their own core designs. I would avoid cameras that use either FLIR Lepton or Seek Thermal cores as these are showing their age now. For information, TOPDON use the Infiray Tiny1-C core but frame rate is dictated by the power of the host processor used with the core. Video recording can be a lot to ask of a budget thermal camera and is unlikely to be fully Radiometric video.

[torch]

1. UNi-T UTi260B (UTi690B) - has a following on this forum but I consider it’s build quality marginal. Uses the Infiray Tiny1 C core.
2. Infiray C200 - uses the Infiray Tiny1C core and is a nice camera with good build quality.
3. Guide IR PC210 - uses their own Guide Sensmart TIMO256 core and is a well built camera
4. HTi HT18+ - Unknown which core is used but a decent performer from what I have heard. Not tested one myself.
5. HTi HT19 The next model up from HT18+ with higher resolution.

Thanks for the suggestions. For the record, from what I understand, the Infiray C200, Vevor SC240N and Topdon TC004 are all the same unit under different names. The TC005 adds visible light camera and extends the IR range from 300°C to 550°C and adds 3 more palettes. (Interestingly, they run Linux and someone actually hacked one to play a video game -- Doom, IIRC)

The Guide PC210 is on my list already. It's one that keeps cropping up and I haven't heard anything bad about it yet. Eleshop has an interesting comparison with the Uni-T UTi260. The Uni-T has a bigger screen and higher IP rating, but is heavier, shorter battery life, longer charge time, and worse spectral resolution and range. Price is almost exactly the same.  EDIT: at Eleshop. Can be had considerably cheaper from the Orient.

The HTI19 actually looks less impressive than the HT18+. Ok, it has higher resolution but lower frequency, narrower range. Overall, I wasn't impressed with either of them. Lack of range adjustment, short (2 hour) battery life (yes, it comes with an extra battery, but you have to use the camera to charge it, so what's the point?). But most importantly, instead of an IP rating and drop height, the manual contains several cautions -- notably to avoid bringing it in from the cold unless it is in a sealed bag to prevent internal condensation!

[bap2703]

Cons:
Narrowest spectral range (slightly: 8 - 14 um fs 7.5 - 14 um)

Note that narrowest spectral range is a PRO. Provided the sensitivity is the same.

Manufacturers will want to increase the spectral range to collect more light, so the sensor appears to be more sensitive.
The problem is you start to be in a region where the atmosphere is less and less transparent, so your brightness is now more dependent on range and humidity.

In reality it probably doesn't matter for cheap cameras that won't do long range radiometric measurements.

[torch]

Note that narrowest spectral range is a PRO. Provided the sensitivity is the same.

Now there's an interesting point. None of the spec sheets I've seen list overall sensitivity by name and few mention an F-number. As I understand it, the overall sensitivity is simply the thermal sensitivity * the F-number.

I wonder if one could estimate the F-number based on the published pitch, resolution and focal length? Maybe not with complete accuracy, but sufficient for comparisons?

If I do that, then the estimated overall sensitivity comes out at:

Hikmicro B20 (33um)
Topdon TC005 (41um)
Guide PC210 (46um)
Uni-T UTi260B (52um) 

EDIT: How does frequency play into the equation? Does a lower frequency provide more "exposure" time, so a less-sensitive detector can accumulate an equivalent amount of light as a more sensitive detector running at a higher frequency?

[Fraser]

I found Hikvision build quality to be excellent as you can read here…….

https://www.eevblog.com/forum/thermal-imaging/hikvision-ds-2tp31-series-thermal-camera-teardown-307840/msg3942805/#msg3942805

Fraser

[Fraser]

One thing I should say in relation to budget thermal imaging cameras, such as we are discussing here, is that we should not get too hung up on pure specifications. These are NOT like buying a high performance SLR VL camera, they are thermal cameras offering low resolution, reasonable measurement performance and decent thermal sensitivity. Many such cameras use the same thermal imaging core. The specifications of thermal cameras can be, and often are, manipulated in an attempt to stand out from the crowd. With regard to spectral response, the cores use 12um pixels and a pretty broad spectrum AR coated microbolometer window and lens block. Getting into spectral response at this price point is likely a fools errand. All will be similar and cover the common part of the LWIR spectrum.

These are “bargain” budget thermal cameras and not science grade cameras. The buyer needs to consider their intended use for the camera and whether a particular model stands out as best choice, or maybe multiple models fit the requirement. They would be wise to not focus too much on NETD or spectral response 

Fraser

[torch]

Getting into spectral response at this price point is likely a fools errand. All will be similar and cover the common part of the LWIR spectrum.

Point taken. I read your teardown on the Hikmicro and also your adventures with a Uni-T. Interesting differences, and reflected in the price.

One feature that I am interested in (based on experience with a high end TIC) is manual range and span control for the purpose of fine-tuning the response based on the task at hand. Of the models listed in this thread, only the B20 and the PC210 offer that. What is your take on it?

[Fraser]

I consider manual span and centre temperature an important feature as well. I agree, that is highly desirable 

Fraser

[bap2703]

Agree with Fraser on the specs.
Many are irrelevant for a simple use, while many relevant are never published.

I just bought a TC004.
Main feature: amazon could deliver it right away 
Never published spec: latency... there's something like 200ms between capturing and displaying. That's more annoying than absolute temperature accuracy for me.

[torch]

Epilogue: In the end I bought the PC210. It was the best value for a camera with the features I deemed most important. Oddly, the best price with shipping for me was from "UNI-T Tools Store" on AliExpress. Weird.

Anyway, I've had it a few weeks now. It feels solid and durable, comes in a nice box with hard foam cut-outs (cardboard, but certainly not a flimsy shipping box  and cheap styrofoam) together with a 32GB SanDisk brand microUSB card, USB-C cord, US charger and wrist strap. I bought a close-up lens from a different source, but still have to get a holder 3d printed. I can't say I've used it to the limits of its capabilities just yet, but it has been handy for the things I have used it for -- checking tire temps on my MC, finding a warm brake disk on SWMBO's car before it became a problem, checking for heat loss in my son's abode as he re-insulated, etc.

Two quirks I have noticed. 1. it remembers most menu settings, but not the manual range setting. Not a huge deal. The auto ranging constantly scales from the coldest to the hottest pixels currently displayed. Switching to manual mode freezes the range at the current values, which can then be tweaked as desired. But if one cycles through the auto and span modes, the manual values are reset to the latest range values. That said, if the camera is in manual range mode when powered off, it will power back up in manual mode at the previous range values. The span mode behaves similarly -- the span is set according to whatever the Auto values happened to be, however, as the camera is panned, the range then floats up or down as required while maintaining the span.

The second is an annoying occasional freeze of the display for a second or so. As if the processor is momentarily busy. I've tried various settings like changing the resolution of or even turning off the visual camera but can't see any rhyme or reason to it. Annoying, but not constant and not a deal-breaker.

Overall I'm quite satisfied with the performance, especially considering the price. Very good value for the money, IMHO.

[DaJMasta]

Is there a little sound that accompanies the freeze?  Could be the Flat Field Correction shutter, to keep the thermal drift of the sensor itself from effecting image quality, in which case it's normal and difficult to get around unless there are FFC timing settings available in the menu (some cores have this, it's rare in handheld inspection cameras as it is required for measurements to be accurate.)

[torch]

Mystery solved.

The click is so faint that my old ears can't hear it unless it's held up to my ear. From which I could determine there are 2 double clicks on boot, and an occasional double click thereafter. But I can't hear either if I can see even a corner of the screen. So I held a stethoscope to the side of the camera and yes, the display freeze coincides exactly. It freezes at the first click and resumes on the second. I assume first is the shutter closing and second is the shutter opening.

Thanks!

[rosetyler]

Thank you for this thread.  I read it while trying to pick my first thermal camera. My PC210 from amazon just arrived. I also notice a very small freeze on the display when the shutter sounds.

I am not sure if I am misunderstanding how the manual mode works. Lets say on manual mode if I tweak the range from 60F-70F and place my hand in the frame, I thought my hand would be a color not listed on the right, such as white. But my hand is always the top color of the range, even though it's temperature is a lot higher than 70F.   My PC210 arrived with: Version: V1.4.7.20230425 and Firmware: V1.2.3.20230112  I found this thread which says there is a firmware bug: https://www.eevblog.com/forum/thermal-imaging/guide-pc210-thermal-camera-anyone-familiar-with/msg4128304/#msg4128304 But I don't see images posted of what manual mode should look like and I'm very new to this.   Does your manual mode work like mine does and what firmware version does yours have? I wrote Guide Sensmart on amazon but I haven't heard back.

Thanks again!

[torch]

Mine is apparently an older version: V1.4.5.20230117, but the same firmware: V1.2.3.20230112.

I think yours is working properly, depending on which pallet you selected. On the right is a bar showing the pallet range. Anything above the highest temperature setting will be the top colour. Think of it as a threshold. Similarly, anything below the lowest temperature will be the bottom colour. The difference between the two colours is called the "span". So if, for example, you selected the pallet from black to red, anything equal to or below your 60°F will be black and anything equal to or above your 70°F will be red.

When changing the manual modes, you can select one of 4 arrows. Pressing the OK button when the top arrow is highlighted raises the temperature range (both min and max) on the scale by equal amounts -- ie: without changing the span. Pressing OK on the bottom arrow lowers the temperatures equally. The left and right arrows increase or decrease the span -- that is, the difference between the upper and the lower temperatures by equal but opposite amounts.

[rosetyler]

Torch, thank you so much for your detailed explanation about how the manual mode works. It really helped me to understand. 

I'm very excited to learn how to use this tool more. I really like it so far.

[torch]

I can't say I've used it to the limits of its capabilities just yet, but it has been handy for the things I have used it for -- checking tire temps on my MC, finding a warm brake disk on SWMBO's car before it became a problem, checking for heat loss in my son's abode as he re-insulated, etc.

One somewhat unusual use I have found for it is monitoring material temperatures when machining stainless steel. 304 SS can be readily machined -- IF one stays within a certain envelope. It does best with a minimum tool speed, but is very sensitive to heat and will work harden instantly if it gets too hot. Here I was cutting a very thin but deep slot in a small diameter rod and the camera was perfect for judging the heat buildup:



Here's another, showing the parallax between the IR and visual images at close range:

[the_cake_is_a_lie]

I read another thread and realized I want/need manual level and span. This thread is more active so may as well ask here.

The Guide PC230 can do manual level and span right? Is there a discernable reason why it costs $359 versus the PC210 at $279 before tax? It has <= 50 mK thermal sensitivity versus <= 45 mK for the PC210. I realize the PC230 has a narrow field of view versus wide, with narrow being better for me to use with electronics.

The Hikmicro B20 has manual level and span. Does anything else like the Hikmicro E1L, TOPDON TC004 or HTI HT-18+?

[torch]

The primary difference is field of view.

The PC230 has a field of view of 25° x 19° with an instantaneous field of view of about 0.1° giving it a focal length of 7mm vs the PC210 of 56° x 48° with an IFOV of 0.215° and focal length of 3.2mm. Basically, the PC230 is more like a zoom lens and the PC210 is more like a wide angle lens. The distance-to-spot ratio of the PC230 is 585:1 (ie: can measure a 1" diameter spot at a distance of 585"). The d:s ratio of the PC210 is 266:1, so the PC230 can "see" the same size spot at a little over twice the distance.  At the other end of the scale, the minimum distance of the PC230 (without macro lens) is 1/2 meter, while the PC210 is 0.3 meter.

Using a macro lens for close-up work, the PC230 is reportedly very finicky about distance to keep things in focus. On the flip side, the PC230 has a longer range for viewing distant objects while still being reasonably accurate with a crisp image. In part, the quality of a distant image is due to the autofocus function. The PC210 is fixed focus, which is a bit of a compromise between near and far image quality.

Other than that, the specs are virtually identical.

[TheKid]

I’ve done my research and it looks like the B10 is ~$50 more than PC210. It appears to me out of the two the only major difference is the windows software. Guide software seems to have issues with edits. The edits remove all metadata and are smaller in size. Is the B10 worth the 50+ ?

Maybe I need a post of my own? I have other questions as well.

[the_cake_is_a_lie]

Cool thanks for clearing that up and explaining the jargon! All I'm interested in is up close and personal electronics so a finicky distance to keep things in focus is no dice. I found that the FLIR C3-X (over $500) and C5 (over $800) do manual level and span from their Amazon descriptions. I see the Hikmicro ($270) B1L, ($430, but see $70 Amazon coupon) B10, and ($550) B20 all have manual level and span on their spec sheets. I assume the ($300) B01 can too when Amazon says it's a newer version of the B10 but E1L is auto only.

I emailed the Amazon rep for HT-18+ and HTI-19 who responded the next day to say that they only have auto level and span and that the HTI-19 is discontinued so it's recommended to order the 18+. If you don't need manual mode then the 18+ has a really good $220 price point for 320x240 IR image resolution and 300k pixel visual image resolution. Has a built-in rechargeable 18650 battery and 3GB SD card.

I’ve done my research and it looks like the B10 is ~$50 more than PC210. It appears to me out of the two the only major difference is the windows software. Guide software seems to have issues with edits. The edits remove all metadata and are smaller in size. Is the B10 worth the 50+ ?

Maybe I need a post of my own? I have other questions as well.

PC210 is cheaper in general but check out the $70 coupon on the B10 and $300 B01 looks like the B10 with longer battery life. Not sure why it's cheaper when it seems superior but maybe I'm missing something in the specs.

[torch]

I see the Hikmicro ($270) B1L, ($430, but see $70 Amazon coupon) B10, and ($550) B20 all have manual level and span on their spec sheets. I assume the ($300) B01 can too when Amazon says it's a newer version of the B10 but E1L is auto only.

The B1L is only 160 x 120 resolution. I can't find mention of a "B01" anywhere on the Hikmicro website. The closest I can find there is the "HM-B01-Pouch", which is indeed a storage pouch for their B-series imagers. I could only find the "B01" imager in the Hikmicro Amazon store. I note that the manufacturer's product number for that listing is actually shown as "B01(Amazon)". The listing is sorely lacking in details making me suspicious that it is built down to a price point exclusive to Amazon.

[Henry1983]

The Hikmicro B01 doesn’t appear to have a visual light camera. Deal breaker.