Author Topic: Flir Vue damaged board  (Read 791 times)

0 Members and 1 Guest are viewing this topic.

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Flir Vue damaged board
« on: September 11, 2019, 03:55:27 am »
First time on this forum. I hope I'm doing it right...

I'm integrating a Flir Vue to a drone system and accidentally powered it with 12V (the camera is rated 5V). Obviously, I smoked something inside :palm:. It seems I just fried 3 components on a board that seems to me as a power/interface board where de USB connector goes. I already sent a message to Flir to get technical support but I'm guessing my best (faster, cheaper...) option would be repair it myself.

The thing is, as you may guess, there are no schematics online and I cannot identify the burned components. So, I highly appreciate any info I can get on this camera that would help me fix it. My first idea is, if someone have repaired one or owns one and have opened it before (it's really easy) and can provide a picture of that board so I can identify the components I will highly appreciate it.

I'm attaching a picture of my damaged board. Burned components are those three in the top left corner of the board. Thanks again!
 

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 1540
  • Country: us
    • medpants.com
Re: Flir Vue damaged board
« Reply #1 on: September 11, 2019, 05:36:55 am »
Definitely a chance someone has a pic of the board of a working one, but I'd say almost zero chance Flir will offer a schematic or would even do component level replacement if serviced - they'd just swap the board for a good one.

I think with a little googling the lower part may be manageable to find - devices like this use a coded character sequence to save space, but usually searching something like "smd code ____" will give you some likely hits, then a little reverse engineering can find you your part.  With a little cleaning (isopropyl alcohol on a cotton swab) will probably get you the SMD code of the rightmost part, but the leftmost one may have enough package damage that it's tough to figure.  From the surrounding parts, I'd say the rightmost one is almost certainly a voltage regulator, the lower part is probably a TVS diode or similar (connects to USB signal lines and a big ground trace), and the leftmost one is a little harder to determine - no obvious giant traces or filter caps, at least, though, that bypass cap and thicker trace could suggest a second voltage rail.


Those pads on the bottom could really be useful for troubleshooting - you can probably buzz out what pins of the burned parts go to ground or power, which can give you a clue as to what they do or what pinout they have.  Then there's a chance that by removing a damaged one, something comes good (if it's shorted as its failure).  For example, if you removed all three and replaced the rightmost part (the regulator), there's a chance it would come back up without the other two, or maybe that one of the output options didn't work, but at the very least the voltages on the pads at the bottom would change and give you further clues as to what's going on.


In any case, I'd clean up the damaged parts to see if you can pull any part numbers and maybe someone with a vue can help.  If you find a review of the camera, if there's any shots of the internals they may have an image that could help you out too.
« Last Edit: September 11, 2019, 05:38:26 am by DaJMasta »
 
The following users thanked this post: rodri_e

Offline Cat

  • Contributor
  • Posts: 40
  • Country: de
Re: Flir Vue damaged board
« Reply #2 on: September 11, 2019, 07:17:24 am »
Definitely a chance someone has a pic of the board of a working one
Pictures are available here, unfortunately the marking is not clearly visible:
https://www.eevblog.com/forum/thermal-imaging/death-of-a-camera-by-drowning-!-flir-vue-its-vulnerabilities-and-death/
With the marking code it should be possible to get the part number.

« Last Edit: September 11, 2019, 08:17:27 am by Cat »
On the Internet, nobody knows you're a cat.
 
The following users thanked this post: rodri_e

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #3 on: September 11, 2019, 04:34:57 pm »
DajMasta, thanks for your reply. If Flir replaces the board that would be great. It shouldn't be that expensive since this seems to be just an interface and power board. My problem right now is time. I'm waiting for their response.

About components, I think I already identified one; the left-most one. The code reads CU45 which I found to be a dual inverter gate, probably for some data interface (though not documented in the user manual). The code of the lower one can be half-read. Seems to me its a 4-number code that starts with 50 (5035?) but so far I haven't been able to find something similar with that code. I'll search a bit more this afternoon. The right-most one is the most damaged and I can identify just the number 3, which can be the first or the last of the code, but I agree it must be voltage regulator.

I already tried buzzing the pads but as you said most are short-circuted with ground. Actually PWR and GND are shorted. I'll do as you say and remove the fried components and buzz the board again to see what I can find.

Once again, thanks a lot for your input and tips. I'll report back as soon as I can work on the board.
 

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #4 on: September 11, 2019, 04:42:08 pm »
Pictures are available here, unfortunately the marking is not clearly visible:
https://www.eevblog.com/forum/thermal-imaging/death-of-a-camera-by-drowning-!-flir-vue-its-vulnerabilities-and-death/

Cat, thanks for the link. I've already checked the pictures there. In fact, that post was the reason I came to this forum for help. I already contacted Fraser (the owner of the thread) and he told me those are form an ebay auction and that he doesn't have full resolution pictures. He advised me to ask for help in this forum and here I am.
 

Offline railrun

  • Regular Contributor
  • *
  • Posts: 72
Re: Flir Vue damaged board
« Reply #5 on: September 11, 2019, 04:54:30 pm »
FLIR and spare parts is pretty difficult.
I can upload some pictures when I get home from my business-trip, but not before this weekend. I only have pictures from other PCBs on my iPhone.
 
The following users thanked this post: rodri_e

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #6 on: September 11, 2019, 06:49:54 pm »
I can upload some pictures when I get home from my business-trip, but not before this weekend.

Excelent railrun! That would be very helpful. The weekend is already three days away  :D
 

Offline ir.ukrm

  • Contributor
  • Posts: 24
  • Country: ua
Re: Flir Vue damaged board
« Reply #7 on: September 13, 2019, 07:00:55 am »
I hope you can see
 
The following users thanked this post: rodri_e

Offline frenky

  • Supporter
  • ****
  • Posts: 966
  • Country: si
    • Frenki.net
Re: Flir Vue damaged board
« Reply #8 on: September 13, 2019, 08:11:40 am »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #9 on: September 13, 2019, 11:21:50 am »
Sadly, in my experience, some of the SMD markings on FLIR core components can be hard/impossible to find via the normal Google searches and SMD cross reference lists :(

In such cases I have had to reverse engineer the area of the PCB surrounding the mystery component and worked out its function and a suitable replacement. With power and I/O related PCB’s this is usually reasonably straight forward. You would expect to see TVS diodes, Linear Low Dropout regulators and DC-DC converters on power input boards. The PCB that is the subject of this thread does not hade the tell tale inductor used in DC-DC converters so there is likely a Low Dropout Regulator and Protection in the form of TVS diodes. I note that the square chip appears to be on a power rail. I fear that chip may not have survived the over voltage event. The fact that most other semiconductors on the PCB visibly fried does not bode well for it :( Note also that I see evidence of a PCB power plane overheating and damage. I suspect the source of 12V was not current limited :( I think a replacement power/ I/O board may be needed. FLIR will likely agree to repair it but expect a bill of around £400 or $400 in the USA. On the up side, FLIR can test the rest of the core to determine if further damage has been done. In general, the fixed price repair covers everything except the Microbolometer.

Your alternative approach could be to reverse engineer the PCB to determine what the various sections of it do and completely rebuild it. After first repairing the physical damage to the PCB power plane. The PCB May have an internal power rail short as I can see at least one feed through that looks to have been destroyed in a very fiery manner ! You can check the TAU2 core interface connector pinout in the TAU2 Manual and see what this power board needs to provide to it. You can also investigate the data and video paths from the core through the I/O part of the PCB. From memory, the VUE is just a TAU core with a power and I/O PCB added to its rear. If you make up a test cable using the damaged PCB’s connector connections to the core you can likely test the core for damage and even talk to it via RS232 and see the video output. Just thinking about it, the larger chip on the damaged PCB is likely part of the USB conversion to talk to the core via RS232. If the core is not damaged, you can work on rebuilding a working power and I/O board. Before using the Damaged PCB, please remove the visibly damaged components and test it for power rail shorts as these could short any voltage applied to the connector when testing the rest of the core. Only use the damaged board as a convenient test connector to connect to the core.

Be aware that if you need someone else to carry out the repair on this VUE, it is going to be expensive due to the time needed to diagnose and repair what is damaged. The “return to FLIR” option can be a good value solution where there is potential extensive damage but a working Microbolometer. FLIR can change out the boards easily and the VUE will be as good as new. This will not be the first time that someone has overvolted a VUE, I can assure you of that.

Fraser
« Last Edit: September 13, 2019, 05:27:24 pm by Fraser »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #10 on: September 13, 2019, 07:34:44 pm »
I had a think about the possible candidates for the USB to UART bridge chip on the Damaged board. I believe it will be a Silicon Labs CP2102N or similar from the CP21xx range. It certainly visually fits the one on the PCB and I know for a fact that the TAU series official VPC board uses the Silicon Labs CP21xx series bridge. Thankfully this chip is inexpensive and common.

https://www.flir.co.uk/support-center/oem/vue-camera-controller-gui/

Fraser
« Last Edit: September 13, 2019, 07:38:59 pm by Fraser »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #11 on: September 13, 2019, 07:58:41 pm »
Just looked at the pinout of the FLIR VUE and from my knowledge of the TAU and the I/O on the VUE 10 pin connector, I would expect to find the following chips present on the I/O PCB.

1. Power input protection using TVS diodes
2. Low dropout 5V regulator for regulating the supply to the core
3. USB to UART bridge for core serial communications (CP21xx series)
4. Video buffer to provide isolation of the cores video output.
5. There may be a logic buffer between the USB-UART bridge and the core UART I/O

Fraser
« Last Edit: September 13, 2019, 08:11:21 pm by Fraser »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #12 on: September 13, 2019, 08:45:55 pm »
In case it helps..... how I would proceed with this ‘patient’.....

1. I have a VPC board that I built myself and I would connect it to the core to test for damage to it. I can communicate with the core via USB as my DIY VPC contains a CP2102 USB-UART bridge. The Standard TAU GUI may be used to talk to the core.

See details of my DIY VPC in this thread......

https://www.eevblog.com/forum/thermal-imaging/thermal-camera-teardown-the-flir-tau-320-by-fraser/

2. The damaged power and I/O PCB is showing a short on the power rail. This could be in the TVS diode, LDO regulator or an inter power plane short in the PCB itself. The visibly damaged components should be carefully removed and the PCB power plane checked for shorts. Hopefully none will be present but parts of the power plane may have gone open circuit if any feed through shave burnt away. There is pcb overheating damage clearly visible near to the LDO regulator. Remove any and all power rails shorts by whatever means is best.

3. Once the damaged PCB no longer presents a short across the power rails, and all damaged components have been removed, it is time to carry out continuity checks from the known pinout of the 10 pin connector to the various damaged component pads. This will identify the purpose of the components. It would also be good to carry out continuity from the TAU core connector to the damaged components, if any.

4. Once the power rails have been clearly identified I would apply 5V current limited to 500mA to just the power input of the board (without the core connected) and then to the output pin pad of the LDO regulator. There should not be an over current situation. If there is an over current situation, another component has failed. Check the USB-UART for self heating as it may have been damaged if the LDO regulator went short, input to output. If no over current situation occurs we can move on.

5. Connector the damaged PCB to the core and apply a well regulated 5V current limited power supply to the output pin pad of the LDO Regulator. Set the current limit to 1A. Switch on the power to the PCB and core. Listen for the FFC shutter operating. If it does. The core has completed boot which is very good news ! If the FFC shutter does not operate, the core may have suffered damage.

6. If the FFC shutter can be heard, connect the USB control cable and see if the VUE cab be accessed by the GUI. This assumes that one of the removed components was not in the USB data path ! Hopefully not.

7. If the FFC shutter is working, use a monitor plus test lead or oscilloscope to probe the video lines in the damaged PCB. If there is a video buffer chip and it was one of the chops removed, you will need to probe on the Core connector side of its pads. Hopefully video will be present as the buffer is not essential and is just a protection stage for the core.

I will stop here as that is enough to be going on with for now.

Fraser
« Last Edit: September 13, 2019, 08:49:36 pm by Fraser »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #13 on: September 13, 2019, 10:00:46 pm »
Now some component data.......

The 4 pin “503B” is a SP0503BAHT Unidirectional TVS array.

Datasheet is here :

https://www.mouser.com/pdfdocs/Littelfuse_TVS_Diode_Array_SPA_SP05.pdf
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #14 on: September 13, 2019, 10:40:07 pm »
The SMD component marked “HR4nC” is a 5V Low Dropout Linear Regulator. It’s current rating must take account of the needs of the core. The core usually pulls around 450mA but peak current during an FFC event is 0.8A. I would fit a 1A rated component, if available. You need an SOT23-5 format part for a perfect fit. A lower current regulator may work but I would consider 0.5A the absolute minimum to risk otherwise poor regulation or even regulator shutdown may become an issue. 1A would be best.

The pinout of the regulator on the damaged PCB looks standard. I have marked the input, output and 0V pads for you.

The HR4nC is present in other FLIR cameras that I have worked on. I believe it is a custom ID as I have had no success in tracking down any data on it. When all is said and done, it is just a 5V LDO Regulator in SOT23-5 case format. I thought it was a Microchip MCP1802 but that is only rated at 300mA so not a match.
More than likely it is a Linear Technologies or TI device.

Fraser
« Last Edit: September 13, 2019, 10:41:39 pm by Fraser »
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #15 on: September 13, 2019, 11:01:17 pm »
Now the bad news  :(

Having revisited your picture of the damaged PCB, I would suggest the following failure sequence and it is not a happy read  :(

1. 12V is applied to the camera that has a 6V maximum input rating.

2. The HR4nC LDO Regulator maintains 5V at its output for a period of time but is being over-stressed.

3. The HR4nC LDO regulator fails input to output and 12V appears on the PCB’s 5V supply rail. This power rail supplies the USB-UART bridge,  CU45 and camera core so is fitted with TVS diode Protection.

4. The 503B TVS diode array is activated and clamps the 5V supply rail in an attempt to prevent damage to components connected to it. In the face of a non current limited constant 12V supply it quickly overheats and succumbs to the event. It may have failed short circuit as it died but the short would cause the PCB power plane to heat up and potentially burn the PCB at points of highest resistance. The TVS or PCB might actually burn out to an open circuit, given time.

5. The CU45 single inverter gate has been destroyed. This IC sits on the TVS protected 5V supply rail yet it has been destroyed in a pretty aggressive manner from the blistered casing. This suggests that the 5V supply rail reached a significant over voltage condition for long enough to boil the CU45. This does not look good for any other sensitive components and IC’s that share that 5V supply rail.

From the above, there is a good chance that the USB to UART bridge and core have experienced at least a brief overvoltage on their 5V supply rail. Whether damage occurred is dependant upon the maximum input voltage tolerance of those components. If not capable of withstanding 12V with high high current capability for as long as it took for a short circuit to form across the supply rails, then serious damage will have resulted.

Fraser
« Last Edit: September 14, 2019, 02:22:18 am by Fraser »
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #16 on: September 13, 2019, 11:09:33 pm »
The CP2102 USB to UART bridge that I believe is used on the VUE camera is rated for a Maximum of 5.8V on the VBUS pin. Sadly I believe that VBUS pin was exposed to more like 12V so damage is likely. Plan to replace the CP2102.

Datasheet is here......

https://www.silabs.com/documents/public/data-sheets/CP2102-9.pdf

Fraser
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #17 on: September 13, 2019, 11:17:14 pm »
I believe that I have identified the composite video PCB traces so it is easy to exclude the video filters etc from the reverse engineering of the PCB. I have marked them in Red on the picture. There are the two large annotated Video output pads for testing purposes. There is no video buffer amplifier on the damaged board.

Fraser
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #18 on: September 13, 2019, 11:24:34 pm »
The USB inputs and UART outputs of the USB to UART bridge chip are shown on the attached image.

Fraser
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #19 on: September 13, 2019, 11:42:29 pm »
From my past experience with the TAU core I think I know the purpose of the CU45 inverter gates. The output from a CP2102 is standard UART Logic ....  5V = Mark and 0V = Space

The TAU uses inverted UART  Logic of 3V3 = Space and 0V = Mark. As such an inverter gate is required between the CP2102 I/O and the TAU I/O. The TAU2 has auto data polarity detection and correction. It still boots with inverted UART data polarity however.

The head scratcher for me at the moment is that the TAU uses 3V3/0V data levels so the CU45 running on 5V would overvolt the data rails on the TAU. This is normally addressed by supplying the CU45 from a 3V3 rails so it acts as an data inverter level changer.

It is late here in the UK so I will leave this here for tonight. Many thoughts are going through my mind.... is the LDO regulator a 3V3 device intended only for the power needs of the Power/I/O board and the TAU directly connected to the cameras 5V input ? That is an ugly possibility as the SOT23-5 format is not normally a high enough current LDO regulator for a whole camera. Oh dear, as I write this, a picture is forming :(

Could it be that the VUE Camera cable 5V is fed straight to the TAU power supply DC-DC converters and the CP2102 and CU45 have their own low current 3V3 Regulator on the power/I/O board ?

Time to do some continuity checks on that power board and see if this is the case. Then time to build a test rig to test the TAU core as it may well be dead due to the overvoltage :(

Fraser

Fraser
« Last Edit: September 13, 2019, 11:45:58 pm by Fraser »
 
The following users thanked this post: rodri_e

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #20 on: September 14, 2019, 12:09:53 am »
The 503B TVS array is involved in protecting the LDO regulator output supply rail and the USB D- and D+ lines. I have shown the PCB traces for USB on the attached image. They go up to the TVS diode array and then on to the CP2102 USB input port.

Fraser
 
The following users thanked this post: rodri_e

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #21 on: September 14, 2019, 12:18:37 am »
I hope you can see
Very clearly!! Thanks A LOT!!
 

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #22 on: September 14, 2019, 12:57:16 am »
Hey Fraser, you almost fixed it for me!! I didn't read this thread again before starting to remove components and buzz paths and came about the same conclusions you did. More or less.

First of all, as soon as I removed the components, the short in the power rail was gone. Then I was able to look and buzz around the board. You're right about the TVS and the inverter gate. Also that the HR4nC must be a regulator but the code can't be found with google. Also you're right about VBUS directly connected to PWR but also VREGIN (by sharing the same path) and VDD (it buzzes... I don't know if it's intended or some component shorted to PWD). I don't really understand this configuration because it doesn't fit into what is suggested in the data sheet.

Now the good news (I hope!). There is no direct connection between PWD and the TAU core connector. There is what seems to be a separate power rail after the HR4nC that do buzz with it. You can see it above the CP2102 with two big capacitors to ground. There's actually a small led there that I guess should lit when the camera is powered (I found it cause it lit up when buzzing between this regulated power line and ground). So, as you initially stated, the LDO is powering the TAU. The burned HR4nC may have protected the module.

On the other hand, it's very likely that the CP2102 it's also damaged from overvoltage but, since I don't need data interface with the camera, just video output, I guess I can just ignore it and also the inverter gate. Am I right?

Next test would be what you suggested yesterday, apply regulated voltage to the power line after the HR4nC. I haven't done it because I'm not sure if that is really a 5V or 3V3. Giving the fact that the module is power by this LDO... should it be then 5V, correct?.

Thank you very VERY much for your help. Now it seems easy to fix but it's the years of your experience posted here what makes it look easy.
 

Offline rodri_e

  • Contributor
  • Posts: 10
  • Country: co
Re: Flir Vue damaged board
« Reply #23 on: September 14, 2019, 01:30:40 am »
The pinout of the regulator on the damaged PCB looks standard. I have marked the input, output and 0V pads for you.

I haven't checked the images you uploaded until now. We agree on everything except that in this one what you've marked as output is connected to PWR, so it should be input, and what you've marked as input is connected to the TAU connector, so it should be output. Right?
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 8415
  • Country: gb
Re: Flir Vue damaged board
« Reply #24 on: September 14, 2019, 10:09:14 am »
Hi,

I am working without ‘hands on’ and only from your picture so please do consider my comments in that light and make your own decisions based upon the readings you are seeing in tests. I thought the LDO was a standard pinout but there is the possibility it is not. To be certain I would check continuity between the power input cable and the pads of the regulator. Whichever pad has continuity with the power cable must be the input pin :) I am uncertain of the function of those large annotated video and power pads on the PCB. There is the possibility that they are both outputs to accessory boards with the Power pad actually a supply to accessory boards rather than an input to the camera. I honestly do not know.

The TAU2 core operates on 5V but its data communications operate on 3v3 logic levels, hence my comments on the level conversion needed for the USB-UART bridge output to the core.

It would be really useful if you can connect a regulated 5V to the supply rail of the core using the pads on the damaged board. Then monitor the video output. The core is autonomous so should just start and produce an image. If it does, we can breath a sigh of relief as the core survived the overvoltage incident :)

The CP2102 and CU45 are not needed if you are not interested in controlling or configuring the TAU core. Those components are not expensive though so I would fit replacements just in case configuration is ever needed.

All the Best

Fraser
« Last Edit: September 14, 2019, 10:18:16 am by Fraser »
 
The following users thanked this post: rodri_e


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf