Flir E4 Thermal imaging camera teardown

[manu]

But, how do you manage to dare opening quite costly stuff without being feared of failing?

Feel the fear and do it anyway 

My first thought every afternoon when I wake up lol

The true question is: "where do you get these equipments for tearing down?". I got one of these cameras as a loan, but I would not tear it down regarding the caution the distributor was asking.

[olsenn]

http://www.filedropper.com/flire4

Here's an archive of all the FLIR Ex information I could find. May be useful having it all in one location.

It has firmware v1.18.7 as well, so perhaps you can downgrade if FLIR does ship your camera with firmware that blocks the hack

[olsenn]

http://www.filedropper.com/flire4

Here's an archive of all the FLIR Ex information I could find. May be useful having it all in one location.

It has firmware v1.18.7 as well, so perhaps you can downgrade if FLIR does ship your camera with firmware that blocks the hack

In edlens.asp, could these values tell you the part number and serial of the lens to see if the E4 and E8 are different?

   

         restreeLens.setResourceValue( "version", Request.Form( "lensversion" ) );
                restreeLens.setResourceValue( "SN", Request.Form( "lensserial" ) );
                restreeLens.setResourceValue( "PN", Request.Form( "lenspart" ) );
                restreeLens.setResourceValue( "focLen", Request.Form( "lensfoclen" ) );
                restreeLens.setResourceValue( "fNumber", Request.Form( "lensfnumber" ) );
                restreeLens.setResourceValue( "clearanceDistance", Request.Form( "lenscleardist" ) );
                restreeLens.setResourceValue( "rotMirr", Request.Form( "lensrotmirr" ) );
                restreeLens.setResourceValue( "opticalMinMeasRadius", Request.Form( "lensminmeasrad" ) );

[Pinkus]

In edlens.asp, could these values tell you the part number and serial of the lens to see if the E4 and E8 are different?

   

         restreeLens.setResourceValue( "version", Request.Form( "lensversion" ) );
                restreeLens.setResourceValue( "SN", Request.Form( "lensserial" ) );
                restreeLens.setResourceValue( "PN", Request.Form( "lenspart" ) );
                restreeLens.setResourceValue( "focLen", Request.Form( "lensfoclen" ) );
                restreeLens.setResourceValue( "fNumber", Request.Form( "lensfnumber" ) );
                restreeLens.setResourceValue( "clearanceDistance", Request.Form( "lenscleardist" ) );
                restreeLens.setResourceValue( "rotMirr", Request.Form( "lensrotmirr" ) );
                restreeLens.setResourceValue( "opticalMinMeasRadius", Request.Form( "lensminmeasrad" ) );

I would guess this is as the software is beeing used for several Flir models, even the ones with exchangable lenses.
I looks like the lenses of E4 / E8 are identical. At least the size is identical but the material might differ. Need to do the same shot at the same time with both cameras and then a comparison of the two pictures.

[mrflibble]

Need to do the same shot at the same time with both cameras and then a comparison of the two pictures.

And then swap out the lenses and do another comparison.

[Pinkus]

Need to do the same shot at the same time with both cameras and then a comparison of the two pictures.

And then swap out the lenses and do another comparison.

If there are differences seen - yes, then a lens swap should be made.... no need to change them back

[Pinkus]

Thank you for the trick!
But, how do you manage to dare opening quite costly optics stuff without being feared of failing?

I am sure Mike, as most of us, did open new devices every since he can remember. Probably has something to do with genes. Thats why we are here on EEVblog and not here: http://www.pinterest.com/julia/mens-fashion/ 

[mrflibble]

If there are no differences seen, then yes a swap should be made. It just depends on how systematic you want to test. But hey, you're the lucky bastard with both an E4 and an E8, so you get to decide the test method. 

[mrflibble]

I am sure Mike, as most of us, did open new devices every since he can remember. Probably has something to do with genes. Thats why we are here on EEVblog and not here: http://www.pinterest.com/julia/mens-fashion/ 

That website sucks. It doesn't include a "doesn't require ironing" score, so how on earth are you supposed to make an informed decision? 

[Taucher]

Some Ex0 model would be nice to check what's beeing used as WLAN / BT
I wouldn't be surpised if the double-digit series cameras would be built around the same core design - display and sensor specs look identical to me.
(in my imagination it's even a similar/same/"differently bonded" core as used in the Quark products - but note 17µ(640x) vs. 25µ(320x) difference in tech-spec)

It's horribly expensive to produce a chip just for one small batch of products - the more sold units can use the same components, the better... just look at the date of the sensor-die... clearly a modern design just fabricated for the E-series ... I'm dead sure that applies to the default lens as well!

Where the E8 model really differs: you get an additional battery and a nice charger for the couple extra grands....

[mrflibble]

Some Ex0 model would be nice to check what's beeing used as WLAN / BT
I wouldn't be surpised if the double-digit series cameras would be built around the same core design - display and sensor specs look identical to me.

Wouldn't surprise me either. With a bit of luck they use it as bluetooth uart with bspserial. If so, then even if you couldn't get the exact same module you would probably still be able to rig up something decent. That said, it would be really nice if someone could check what bluetooth module is being used in the higher end models.

Where the E8 model really differs: you get an additional battery and a nice charger for the couple extra grands....

Does it include one of those oxygen free usb cables?

[Pinkus]

If there are no differences seen, then yes a swap should be made. It just depends on how systematic you want to test. But hey, you're the lucky bastard with both an E4 and an E8, so you get to decide the test method. 

No, I do not have an E8. I thought, at a company I know, they would have an E8 but I learned on Monday it is only an E6. I tried to lend it for a day, but they did not want to give it away 
Maybe my contact just had a bad day. I will try this again in a week or so.

[Taucher]

Just some notes regarding "Exportable framerates" (taken from simlar specsheet):
7.5 Hz (NTSC 30/60Hz) = factor 60Hz div 8
8.3 Hz (PAL 25/50Hz) = factor 50Hz div 6

EDIT: Measured update rate@display: 9 Hz

... and my E4 has just arrived!

Code: [Select]

.version.swcombination.fullver text "os16.0.10_appkit1.0.13_prodkit1.0.0.7"
.version.swcombination.ver text "1.18.7"
@64bit drivers: they auto-add windows-firewall-rules to allow inbound connections to flir processes

Update: mod successful
@Comparison.png: images blown out of scale by my software

[solarbot]

TOTALLY AWESOME - thanks Mike.

Worked on my E4.10 with Firmware 1.19.8

[eliocor]

Just ordered a brand new E4:
total amount (taxes+s/h included) = 978 Euros.

[Fraser]

Regarding differences between the E4 and E8. They may use the same microbolometer and lens, but Mike has highlighted the presently unexplained doubling of thermal differential resolution in the E8. I recall that during Mike's OS investigations he saw two microbolometer modules listed. It is possible that the higher specification units do have a different, higher spec, Micro-bolometer fitted.

I must say that the 0.06 degrees Celcius E8 specification is impressive for a relatively cheap unit with an uncooled Micro-bolometer detector, as opposed to a cryo cooled semiconductor type. My best (megabux) FLIR uncooled industrial camera is spec'd at 0.08 Degrees C  thermal differential resolution. I may be wrong here, but IMHO lens size sets the sensitivity of the camera system, but would not influence the differential Temp detection capability of the detector.

Either FLIR hobble the E4 and E5 units to lower temperature resolution or a different detector is used in the E6 and E8 units.

At the end of the day, 0.15 Degrees Celcius temperature resolution is very respectable in a sub $1000 unit, and is the same as my Circa 1997-2000 $40k FLIR PM570. The PM570 is not considered insensitive to temperature differential in industry and has served well over the years. More modern cameras from FLIR acieve far greater thermal resolution (~30mK) but at significant retail cost.

For info:

FLIR PM570 - 1st Generation uncooled micro-bolometer - Differential temerature resolution = 0.15 Degrees.
FLIR PM695 - 3rd Generation uncooled micro-bolometer - Differential temperature resolution= 0.08 Degrees.
FLIR P65- 3rd Generation uncooled micro-bolometer - Differential temperature resolution= 0.08 Degrees.
FLIR E40-  ? Generation uncooled micro-bolometer - Differential temperature resolution= 0.07 Degrees.
FLIR E60- ? Generation uncooled micro-bolometer - Differential temperature resolution= 0.05 Degrees.

The lenses of the E40 ans E60 are far smaller than the large aperture lenses fitted on the PM and P series cameras yet they produce better thermal resolution figures. I think anyone looking for better thermal resolution in the E4 should remove the lens from their equation, unless they also desire greater overall system thermal sensitivity, for greater range (distance) operation.




 

[Taucher]

Regarding differences between the E4 and E8. They may use the same microbolometer and lens, but Mike has highlighted the presently unexplained doubling of thermal differential resolution in the E8.

IMHO it's probably much simpler: marketing guy + add noise function for lower end models

[muvideo]

Just ordered a brand new E4:
total amount (taxes+s/h included) = 978 Euros.

Hello, did you buy it in Italy?

Fabio.

[mrflibble]

I must say that the 0.06 degrees Celcius E8 specification is impressive for a relatively cheap unit with an uncooled Micro-bolometer detector, as opposed to a cryo cooled semiconductor type. My best (megabux) FLIR uncooled industrial camera is spec'd at 0.08 Degrees C  thermal differential resolution. I may be wrong here, but IMHO lens size sets the sensitivity of the camera system, but would not influence the differential Temp detection capability of the detector.

I am a total thermal imaging n00b, but the bit of reading I did gave me the impression that ... well, what you said! XD Very nice temperature resolution for this type of sensor. That was another reason I could tell myself I really had to buy this nice bit of gear.

Do you know if for uncooled microbolometers there are any improvements to be had by making sure the sensor has very uniform temperature. Or maybe by cooling the sensor below ambient.

Another way of putting this: do you have any links to reading material so I can RTFM on this type of sensor?

[eliocor]

Yes and no.
I bought it from a VERY respectable vendor: the Italian branch of Omega (but the shipping will be made from UK).
Strangely you can not find the E4 directly, but using the code (found on the German Omega site) OSXL-E4 you can make the order.
The price paid was (shipping from the UK, so VAT was 20%):
800.00 Euros - Flir E4
15.00 Euros - S/H
163.00 Euros - VAT (20%)
------------
978.00 Euros - total amount

The transaction went OK and I have already received the order confirmation*.
Shipping should be done on 2013-11-13 (1 week as defined in the order page)

To my knowledge, here in Italy the default E4 price is 995 Euros (VAT excluded)

*) hoping to not receive a call stating the order is void... 

[Taucher]

OT: ... meanwhile Dave is getting robbed by FedEx for the FLIR review he did ... https://twitter.com/eevblog/status/397976002711658496 

[TopLoser]

Wow..... thanks Eliocor!! I just bought one for 665 GBP plus 9 GBP shipping. And I'm in England, no idea where it's shipping from!

Description           Quantity   Item Value   Item Tax   Item Total   Line Total
OSXL-E4 - FLIR E4   1           665.00           0.00           665.00           665.00
Carriage Charge                                                                   9.00

[bookaboo]

Whaaaat??
Did you order from outside the EU? I suspect there may still be VAT to add to that.

[TopLoser]

Whaaaat??
Did you order from outside the EU? I suspect there may still be VAT to add to that.

I quoted my VAT registration number during the order process, no VAT to pay.

[Fraser]

Mr fibble,

I can assit on your question.

Micro-bolometers are often temperature stabilised using an integrated peltier module. Some newer micro=bolometers have dispensed with the Peltier based temperature stabilisation. I am not sure whether the E series FPA has a Peltier module but I suspect that it has.

There is no advantage in adding additional cooling to the sensor as you will confuse the temperature stabilisation loop. The built in peltier module would try to counteract your external cooling and would draw heaps of current ! Unlike some semiconductor based detectors, there is no need to super cool the micro-bolometer. IIRC from a conversation many years ago with FLIR, teh FPA is stabilised at +30 Degrees Celcius. This may be why the specs often state the thermal resolution at +30 Degrees C.

Take a look at this thread that I produced:

https://www.eevblog.com/forum/projects/thermal-imaging-camera-microbolometer-design-detail/msg322927/#msg322927