new $39 thermal scanner on kickstarter, the deltaK!

[exoticelectron]

Hi Everyone!

I am the designer and manufacturer of a new thermal scanning tool for electronics troubleshooting, the deltaK thermal scanner!

The deltaK design is based on a tool I made and used for years in my work as an EE. It indicates temperature in real-time using a continuous tone whose pitch varies according to the current temperature being scanned versus the starting point of the scan. This makes it very easy to scan a circuit board for hot parts (or cold parts!). The closer you get to the board and the smaller the motions you make the tighter the precision gets. Its also very sensitive to temperature and can even locate studs in a wall based on the slight temperature difference, in some situations.

Its currently on kickstarter and needs about 800 more pledges in 25 days. The price is a special $39 + $3 shipping to USA for the first 100 pledges. International shipping outside the USA is $15. Please help make this product a reality! I really believe in it and think it has many more uses besides just electronics troubleshooting, but it needs to get out there first!

The kickstarter page has videos and more descriptions, check it out! I am using a google link shortener so I can see how many people take a look, its my job to promote it and I am learning so please don't think I'm tracking you or what not! I am just one person doing this whole project!





Click here to see the deltaK thermal scanner kickstarter campaign!----> goo.gl/Ckk8ef

[frenky]

I can see that a lot of effort went into designing this and I respect that.

But to me one thing is missing for this to be useful... a tiny lcd to display actual temperature.

So you quickly scan the pcb or car engine for hotspots with tone pitch and that is great. (If you can't afford cheap smartphone thermal module.)
But when you find a hotspot you will usually want to know how hot it is.
Information if the component on pcb is at 50*C (122*F) or at 150*C (302*F) makes a lot of difference when troubleshooting.

[frenky]

And doing updates for "Backers only" is the worst think you can do on Kickstarter.

All potential backers will think that you are trying to hide something...

[EEVblog]

I can see that a lot of effort went into designing this and I respect that.
But to me one thing is missing for this to be useful... a tiny lcd to display actual temperature.

Or a speech module to read out the max temperature.

[Chanc3]

I can see that a lot of effort went into designing this and I respect that.
But to me one thing is missing for this to be useful... a tiny lcd to display actual temperature.

Or a speech module to read out the max temperature.

Only if it's in your voice!

[b_force]

What is the benefit of using this thing over a regular of the shelf scanner?

(which I barely use anyway because they are not very precise/cumbersome to work with)

Also wonder how long that exposed sensor unit is gonna last, flapping around in the breeze without any protection.

edit:
I find the 'delta K' a little awkward. The standard scientific symbol for temperature is T.
K is only a type of unit you can use for it.
It also suggests that this scanner is noly useful for measuring giving me temperature differences, not the absolute temperature.
(which in essence is also true, but that's not the point)

edit 2:
On the kickstarter website it says the units are in degrees.
Degrees what? Degrees cows? Degrees wood? Or does it measure mechanical degrees?

[ebclr]

Why not this

https://pt.aliexpress.com/item/GM320-Digital-Infrared-Thermometer-Professional-Non-contact-Temperature-Tester-IR-Temperature-Laser-Gun-Device-Range-55/32697751644.html?spm=2114.02010208.3.1.sqols2&ws_ab_test=searchweb0_0,searchweb201602_2_10065_10068_10136_10137_10060_10138_10062_10141_10056_10055_10054_301_10059_124_10099_10103_10102_10096_10144_10052_10053_10142_10107_10050_10143_10051_10526_10529_10084_10083_10080_10082_10081_10110_10111_10112_10113_10114_10037_10517_10078_10079_10077_10073_10070_10122_10123_10124,searchweb201603_1,afswitch_1,ppcSwitch_5,single_sort_0_total_tranpro_desc&btsid=46479967-e97a-40b4-ae28-5d812ccbadf6&algo_expid=13a27828-0e92-49cc-8150-ca9ae260cbfb-0&algo_pvid=13a27828-0e92-49cc-8150-ca9ae260cbfb

Do more for less

[exoticelectron]

Thanks for the feedback guys, to answer a bunch of questions at once:

-Point and shoot IR guns don't scan. So if you have a surface and want to find gradients or hot spots its a very slow cumbersome process of point and shoot.

-A display or voice readout to indicate the temperature when stopped at a point could be useful, but would be alot more expensive at low quantities. Maybe someday. However, in troubleshooting, I never found this necessary. Scanning a surface and hearing the gradients is alot like using a thermal camera, only in a linear fashion. Its really easy to find the anomalies.

-There aren't really any units. The device measures relative temperature and indicates it with sound. So F or C really only matter if you need to know exactly what the degrees per Hz or dynamic range is, which isn't really important given how the device is used.

[exoticelectron]

On kickstarter you can cancel your ple

And doing updates for "Backers only" is the worst think you can do on Kickstarter.

All potential backers will think that you are trying to hide something...

On kickstarter you can cancel your pledge whenever you want. So if you back something then read the updates and want out you don't lose any money. I keep some of the updates backers only because I specifically want to have discussions with people who have already indicated they want one and are interested in it.

[all_repair]

How about indication through brightness, colour or blinking speed to someone that is tone-dead like me.  I got a low ohm differentiator to catch shorted area also through audio pitch, it is impossible to use for people like me.

[exoticelectron]

How about indication through brightness, colour or blinking speed to someone that is tone-dead like me.  I got a low ohm differentiator to catch shorted area also through audio pitch, it is impossible to use for people like me.

A previous version of the design had two LED's one red, one green, which would indicate if the current temperature was higher or lower than the starting point of the scan. But this wasn't that useful so I took it out.

The sound the deltaK makes during use is alot like a metal detector. A sort of warble that makes a clear low-high-low as you pass over a hot spot. So duplicating that visually (and inexpensively) could probably be done with an LED bargraph that auto-ranges with a time constant to allow for minor drifting of the surface temperature during a scan. This way a hot spot would cause a visual spike that would be easy to see.

But in this keychain size form factor there isn't enough battery to run multiple LED's continuously. So I'd have to figure something else out. Maybe on the next version!

[exoticelectron]

Hey! Kickstarter says someone from eevblog made a pledge! I dont know who it was but thank you very much!!

[PlainName]

A previous version of the design had two LED's one red, one green, which would indicate if the current temperature was higher or lower than the starting point of the scan. But this wasn't that useful so I took it out.

I echo the previous poster - sounds is useless to me (I am deaf). For $39 I would give this a punt if it had some visual indication, but otherwise I would be buying a pile of junk. I am sure hearing people would benefit too (quiet office, warbling toner.... I can see the fisticuffs already!).

But... better than red/green LED, why not point an RGB LED forward, at whatever it is you're scanning. Then you can see, on the actual scanned thing, the relative difference as you pan across it. And it's an almost no-cost option compared to normal twin LED indication. Warbles are so 60s, you know.

If you promised that the device as shipped would have something like that, I will put in my $39 tonight.

Edit: to clarify, have the default temperatue, whatever it is, showing green, then shift to yellow then red as it gets warm. Similarly, shift towards blue as it gets cold. Easy peasy (to program and read).

[mikeselectricstuff]

Many years ago when the early thermopile sensors came out I prototypes something similar to this. At the time I'd not heard of chopper opamps so it was a bit too unstable to be useful so never got round to doing any more with it. Since then, when cheap IR thermometers became common I couldn't work out why nobody was putting audio indication in them as it seemed such an obviously useful addition.

I think a major fail on your video is it doesn't show any actual usage.

Considering how cheap thermal imagers are nowadays, I think you may be a few years too late with this one.

 

[exoticelectron]

Thanks for the feedback guys. The deltaK form factor and certain very specific details about how it functions makes it a great tool for troubleshooting circuit boards, which is what I have used it for for years. Point and shoot IR thermometers with a scan function aren't really good at that particular task, because the output changes too slowly and in a way that makes it very difficult to precisely scan a large surface quickly and zero in on an anomaly. Also the form factor of those tools isnt really appropriate for scanning circuit boards on the fly in a production environment or ad-hoc out of the pocket, they are too bulky and clumsy.

Thermal adapters for smartphones are great, and probably do a better job than the deltaK in many instances. But the price tag of $150+ and a smartphone capable enough to use with it is too high a price tag for quite a few people, not everyone lives in a first world country with a high powered smartphone in their hand as a given. Its too high a price tag for plenty of tinkerers, technicians, and engineers in the USA, as well. Also, its a delicate tool and not something you necessarily want to risk in certain dirty or dangerous environments, where the deltaK would have no problem.

In my experience the exposed sensor is a benefit and not a liability. Being able to get as close as possible to the surface you are scanning improves precision. I have not had a sensor break on me and I've been quite a bit less than careful with the prototype version I've been using for years. If you want to compare, consider how fragile your $600 smartphone LCD and $150 FLIR adapter are. The sensor itself is only about $6 and anyone wanting to replace it could very easily if need be.

The video at the bottom of the campaign shows an earlier prototype in use. The functionality is the same and the production version in the smaller case will use the same circuitry with very minor changes.

Heres the video:

https://youtu.be/hCPXpIp4Vq8

[mwalker]

The demo video above sold me, as @mikeselectricstuff said, it would be great if you had some actual usage in the main video.

I really like @dunkemhigh's suggestion with the coloured LED pointing at the subject too.

[exoticelectron]

The demo video above sold me, as @mikeselectricstuff said, it would be great if you had some actual usage in the main video.

I really like @dunkemhigh's suggestion with the coloured LED pointing at the subject too.

Thanks. Ive definitely toyed with the LED idea before. Its not trivial to do though in the keychain sized adapter. Button cells do not like having alot of current drawn from them and its a tight current budget as is. I dont think I can pull that off for this iteration. Theres just not enough time/budget to get back into the design at that level. And I wouldnt want to create an LED output that didnt work very well and provide the same sort of spike detection that the audible output does so easily.

I agree about the video..its alot easier for me (at the moment anyway) to make a professional/clean looking animation, than it is to make a live-action video demonstrating the device with the same quality. But I think I can pull it off. I'll see what I can do. Sometimes content is more important than clean/professional looking.

[Cloud]

Thanks for the feedback guys, to answer a bunch of questions at once:

-Point and shoot IR guns don't scan. So if you have a surface and want to find gradients or hot spots its a very slow cumbersome process of point and shoot.

I don't know where did you get that info but my GM900 (and I assume other from series) has scan function. If you hold trigger button you are in the scan mode instead of hold.
I am sorry, but compared to what you get for 20eur from China and what for 40$ from you.... I would maybe buy it if it would be max 10$

[exoticelectron]

Thanks for the feedback guys, to answer a bunch of questions at once:

-Point and shoot IR guns don't scan. So if you have a surface and want to find gradients or hot spots its a very slow cumbersome process of point and shoot.

I don't know where did you get that info but my GM900 (and I assume other from series) has scan function. If you hold trigger button you are in the scan mode instead of hold.
I am sorry, but compared to what you get for 20eur from China and what for 40$ from you.... I would maybe buy it if it would be max 10$

Why would you buy it at all if your GM900 does the same job?

Check the specs for your point and shoot. Its "scan" mode is really nothing more than repeated, averaged measurements, with something like a 500 msec time constant. Useless for precision scanning something like a circuit board for local anomalies.

The deltaK has a response time at the limit of the sensor, on the order of 30mS, more than 10 times faster. This makes quickly scanning a surface much faster and more precise.

Even if the point and shoot had as fast as a response rate, reading a rapidly changing numeric display while you are moving the device over a surface is extremely awkward and imprecise.

Also, the point and shoot guns put the sensor too far away from the surface. You'll never get the spatial precision to distinguish between components. And the enclosures of most point and shoots are too bulky to be used inside housings or tight areas in operating devices.

As far as displays go, as nice as it would be to have a display showing absolute temperature, its not really what the deltaK is for. Its meant to scan for anomalies. Absolute temperature and numeric readouts would be useful but are too expensive to incorporate at small quantities and keep the price down.

[timb]

I'm in for one. I think it's a neat idea.

[exoticelectron]

Okay guys, I'm not happy with the price of this tool. I think a professional could justify paying $39 for it, but I would really like the price to be even lower. I'm planning on doing a relaunch of the kickstarter at a lower price, and so its time to do some work.

The circuit that the current design uses is a very old one that I came up with many years ago. Its a good circuit, sort of a mixed signal affair with discrete logic, an R2R (discrete!) DAC, and some opamps, which does the job, but its relatively expensive and uses alot of PCB real estate. Its by no means a kludge and I carefully engineered it long ago, but to really push the price of this device down I am going to have to go back to the drawing board. I'm proud of my old circuit but its not helping get this tool into the hands of trillions (yes trillions) of people so off with its head!!!!

In the past I tried a digital thermopile, Melexis makes one. I did not like how it performed. I didnt take good notes..but the prototype I made using it did not perform as well as the original design. I think it was a combination of the granularity of the temperature sensors digital output creating a stairstep in the audio tone which made it useless at detecting small temperature changes.

So my new attempt will be the next logical combo. An analog thermopile to an ADC on a microcontroller. This opens up all kinds of interesting possibilities, since you can do lots of stuff when that data is in a processor.

To kludge a proof of concept together I first used a microcontroller from a totally unrelated project with a very high quality thermocouple A/D connected to a PIC MCU. I connected the analog thermopile in place of the thermocouple and did some coding. It worked great! But that just proved it was possible with great (and expensive) hardware. So to get things real I kludged up another proof of concept. This time I connected the thermopile directly to the humble 10-bit A/D of a 16F616 PIC. Well, not directly, in between I put a precision AD8604 opamp with a gain of 470 and some RC filters. Some coding later and voila! It works very well, just as good as my ancient mixed signal design. And it should be much cheaper!

At very drastic and fast increases in temperature, like dropping a soldering iron in front of the sensor quickly, you can hear a stairstep, but that shouldnt matter in usage because scanning something that hot so quickly would not be a situation you would need to hear very subtle increases in temperature.

At very low changes in temperature, like a thumbprint on cardboard (see video), you can still hear the anomaly easily and smoothly.

Also, since the data is now in an MCU, that has multiple A/D channels, it opens up the possibility of making the much-requested absolute temperature measurements. But thats on the shelf for now. I'm going to properly redraw the schematic and get PCB layout going to see where we stand as far as real production costs.

[mikeselectricstuff]

Okay guys, I'm not happy with the price of this tool. I think a professional could justify paying $39 for it, but I would really like the price to be even lower. I'm planning on doing a relaunch of the kickstarter at a lower price, and so its time to do some work.

I really can't see a price drop attracting significantly more users, and it increases your risk of losing money.
If I were you I'd forget all about KS, set your sights lower and build a small quantity and get them to users to see how useful it really is, and build from there.
TBH a thermal camera offers so much more functionality that anyone who has a serious need for it is likely to just buy one. 
I'm not convinced there is much of a market between cheap IR thermometers and the cheapest thermal imager.

[exoticelectron]

Okay guys, I'm not happy with the price of this tool. I think a professional could justify paying $39 for it, but I would really like the price to be even lower. I'm planning on doing a relaunch of the kickstarter at a lower price, and so its time to do some work.

I really can't see a price drop attracting significantly more users, and it increases your risk of losing money.
If I were you I'd forget all about KS, set your sights lower and build a small quantity and get them to users to see how useful it really is, and build from there.
TBH a thermal camera offers so much more functionality that anyone who has a serious need for it is likely to just buy one. 
I'm not convinced there is much of a market between cheap IR thermometers and the cheapest thermal imager.

The last thing I'm going to do is forget about kickstarter. Right now there are about 60 people on there pledged at $43 each. If I can drop the manufacturing cost I can get them a working device without meeting the very high KS goal and they would be thrilled, and I'd still be able to pay for my time and production costs. Many of them have pledged for uses totally unrelated to electronics troubleshooting and a poll I made shows that a majority of them think they could get at least 1 other person to buy one at $43 if they had a working example to show them.

Again, not everyone in the world has a $600 smartphone they can attach a $150 thermal camera adapter to. Alot of people could use something that scans thermally but is very inexpensive. Right now there isnt an alternative I am aware of anywhere near this price, and thats before I lower it. 

[PlainName]

I would also suggest not reducing the price. Sometimes things can be too cheap - people expect something good to cost money, and conversely something very cheap must be, well, cheap.

Your product is sufficiently novel to most people that they would think it quite a costly thing. $39 is a steal, and to be honest I was somewhat suspicious when I first saw the headline price and only bothered reading to see who spots the scam first

[mikeselectricstuff]

Again, not everyone in the world has a $600 smartphone they can attach a $150 thermal camera adapter to.

You don't need a $600 phone, a $100 one is just fine
And you can buy standalone imagers for about $200
[/quote]
A lot of people could use something that scans thermally but is very inexpensive. Right now there isnt an alternative I am aware of anywhere near this price, and thats before I lower it.
[/quote]
If there is a market in that price range, lowering it won't make much difference to sales - $39 is well within 'novelty purchase' territory. All that reducing the price will do is make it harder for you to make a viable product, and make people question how useful it is due to lower perceived value.
TBH even at $39 I think you'll struggle to make any money once all your 1-off costs are covered.