Unknown core in a Fluke TI105

[Bobertsawesome]

Tore down a for parts Fluke TI105 I got off eBay. The unit has no LCD (strangely someone stole it, but if it was a parts salvage it makes sense, they go for $400 on the aftermarket). Anyways, saw an older post from 2017 where someone inquired about it with no luck. If I can't bodge my way into it I may try and restore the LCD with something possibly compatible.

[Fraser]

From memory, these cameras use a MIPI LCD display interface. I have a TIS10 fluke camera that suffered a failed LCD and I was not able to identify the display controller. I did look at the boot log of a Ti100 camera though and that gave the LCD display controller details. Sadly if the LCD is missing, the boot log is unlikely to contain that detail on your camera. I will see if I kept a note of the boot log from the Ti100 series camera.

Fraser

[Fraser]

Regarding the ‘core’ in the Fluke cameras….. they do not truly have a “Core” as they use a microbolometer (often DRS) interfaced with a bespoke FLUKE hardware and electronics package. Some FLUKE cameras may be remote controlled from a PC but you sometimes  still need a LCD display to set up the link.

Fraser

[Fraser]

May be of interest

https://www.eevblog.com/forum/thermal-imaging/fluke-tis10-thermal-camera-fraser-has-a-new-patient-)/

https://www.eevblog.com/forum/beginners/replacing-an-unknown-rgb-lcd-panel-with-a-generic-type-advice-needed-please/

[Bobertsawesome]

Thanks, as always you are the expert . Yes, I was surprised in my teardown to find everything labeled Fluke, even the flat flex running to the module for what looks like temperature measuring. Kinda expected a standard module to be used here.

The panel that pops up for replacement is an ET0350A6DM6 (https://www.panelook.com/ET0350A6DM6_EDT_3.5_LCM_parameter_19426.html). No luck on a free datasheet but I'll have to see how much they are charging for it on panelook, may be worth it.

I do have a JTAGulator that I can maybe start probing around the board with. I honestly bought this unit just to rip out the module to use it in a DIY sense, but it seems I may just have to restore the original functionality. Fluke advertises this camera as 160x120, but looking at the sensor window I kinda find it hard to believe it would be that low of resolution. It seems to me that they are maybe software down-sizing the resolution to work on a handheld, but I could be wrong.

[Fraser]

The early FLUKE camera models did use a 3rd part complete thermal core, as they were actually Raytek cameras, but then FLUKE moved to in-house camera builds. The Ti20 contains a BAE SCC500 series core. Buying-in cores gets expensive for the manufacturer but they have to do so if they do not have the in-house expertise to buy just the microbolometer and create the required electronics package and firmware. FLUKE bought companies that had the required knowledge.

I looked for the LCD COG controller identity in my archive but sadly I did not keep a record of the boot log.

Fraser

[Bobertsawesome]

So verifying my business email gave me some free credits so I could download the datasheet for free. Fairly sure they have some tracking number on the PDF so I can't upload the whole thing without risking social credit deduction

Here are the important bits of info from the datasheet. Mainly posting it here if someone else stumbles upon the same issue. Display controller IC is Himax HX8238-A

[Bobertsawesome]

Additional photos for archiving. Side note: looks like the latest date code from this unit was 2016. The thermal/laser/visible module housing has a stack of stickers on it, the latest being 2016.

[Fraser]

Thank you for this very useful information that I will add to my archive to assist others 

Fraser

[Bobertsawesome]

So I've found some other panels from the mfg "EDT" that are very similar but have the 40pin config. Wouldn't be an issue since they use the same driver chip, just needs a breakout.

The only real difference I notice is they are spec'd as having a DCLK max frequency of 10MHz, whereas the OEM states a typ 19.5MHz. Since they are using the exact same driver chip (HX8238-A), isn't this a nill issue or are there actual differences in the LCD TFT technology to allow a faster clock rate? Everything else in the specs are exactly the same, active area, panel actual size, etc.

[Bobertsawesome]

Spoke too soon and answered my own question .

It is just the different operating mode of the driver chip, so theoretically I should have no issues since the Fluke is using the Serial RGB, hence the DCLK matchup. Going to order the ET035009DM6 ($30 on eBay) with some breakout boards and will update as I test it.

[Fraser]

I wondered about the model number but thought it likely a Ti100 series camera.

[Bobertsawesome]

My mistake, I meant TI105. Thanks for catching that.

[Bobertsawesome]

Dropping some patents related to this camera, I didn't find any patents from Fluke for a microbolometer module around the timeframe of this camera. My guess is it goes back to some Raytech patents that will need to be crawled.

Shutterless infrared imager algorithm with drift correction
Shutterless infrared imager algorithm with drift correction (cont.)
Ergonomic configurations for thermal imaging cameras
Thermal imaging camera with infrared lens focus adjustment (Not on the patent label but seems to contain information on the lens assembly and focusing screw)

Also some googling, this looks very familiar doesn't it?  https://en.wikipedia.org/wiki/Microbolometer#/media/File:Microbolometer_internals.jpg

[Fraser]

Good spot on the microbolometer.

Infrared solutions was one of the companies bought by FLUKE to create their thermal imaging camera production capability.

Fraser

[ArsenioDev]

Also some googling, this looks very familiar doesn't it?  https://en.wikipedia.org/wiki/Microbolometer#/media/File:Microbolometer_internals.jpg

Well that's a beautiful image, saved to the personal archive. Thanks!

[Bobertsawesome]

Went through with the new display and did some reverse engineering to verify connections were compatible. Looking at the controller datasheet (HX8238-A), for 8-bit RGB the GG and BB lines are left floating or tied to ground, and the D0-7 lines are routed through the RR0-7 solely. Functionally, everything else is the same. I didn't take the time to try and reverse the LCD flat flex to make sure nothing was going to be a problem as there's nothing I could do to mitigate that on the adapter PCB anyways. But, from my quick look through, all connections seem mfg recommended decoupling components.

I just put in my order through OSHPark, so hopefully, I can report back on its functionality soon. I left the GG & BB lines to be solder jumped to ground manually just in case I missed something weird. The HX8238 datasheet mentioned these should just be tied or floated, but I notice on the OEM LCD pics it seems like some lines are tied high while the rest low, but the image resolution isn't enough for me to be sure. So, just in case, they can be left open. I ordered some flex PCBs as well just for giggles, only marginally more $ and they seem like would fit better for this project. The PCB should accommodate the rubber cutout for both the total cutout space and FPC connector, fitting nicely with nothing on top of each other.

Below are some of my documents used for reversing the FPC and may be of use for other Fluke models & LCDs.

[Fraser]

Excellent work 👍

I own a Ti105 so may be able to provide some help with more detailed images of the original LCD FPC if needed.

Fraser

[Bobertsawesome]

Shouldn't need any more pics, referencing the LCD datasheet I think I got all the pins right. We shall see though