Uni-t UTi260B

[TobyG]

That ugly dab of silastic you'd originally mentioned was my doing, I'd gone and replaced that cell with a fully charged (larger) one, just to make rule that out,
but re-installed the original again, I was gonna order a replacement, but my usual supplier didn't have that particular one.

Should've maybe mentioned that, to begin with.

For context, I hadn't used the camera in quite a while, I wanna say over half a year?
I repair some niche older motorcycle electronics as a side gig and really only use it for finding non-obvious shorts there, beyond that, I barely have a use for the thermal cam, but it sure is very nice for that.



Either way, I am curious to see if the fault will be found and what it ended up being.
For what it's worth, I can follow what you're doing, by the descriptions, even though it's not in my skillset 

[Fraser]

Replacement GC0308 visible light camera core ordered from Aliexpress. I ordered the 35mm version as it will reach the connector without issue. I used the faster shipping option so the camera core is due to arrive by 1st July. A pity that I could not obtain one quicker but I must be patient Work will continue on the camera but if it is the GC0308 that is at fault, we will not see progress until next month when the new one arrives.

Fraser

[Fraser]

I just found a used GC0308 on eBay in France  It cost £12 delivered, which is more than I wanted to pay, but I want to cross that component off my suspects list so that I can focus on what is causing the issue, if it turns out to not be the visible light camera. If it is GC0308 core that is faulty, the new one I have coming from China will still be used as its cable is a better fit for the Uni-T camera.

TobyG, I hope you do not mind me adding to the parts bought (MS621 x 1 & GC0308 x 2) but I have the time available in June, less so in July. If I fix the camera you will only pay for the parts and no labour  If I cannot fix it, you pay nothing and I will reimburse your postage.

Fraser

[TobyG]

As I'd stated in the PMs, no worries, do what you gotta do,
for what it's worth, I'm just glad someone is willing and capable to have a closer look, provided Uni-T wouldn't touch it with a 10 feet pole,
not even when I spelled out the magic words of "paid repair".

[Fraser]

The replacement used GC0380 visible light camera arrived from France today. Sadly after testing it with the Uni-T camera I am still seeing the same fault.

The replacement MS621 lithium cell also arrived today and will be fitted soon. I do not expect that to solve the fault however.

The investigation will continue.

Fraser

[Slartibartfast123]

In your boot logs the GC0308 seems to fail somehow early on:

UNIT login:
[    7.044473] g_webcam gadget: g_webcam ready
[    7.187864] get_device_id:write reg error:reg=fb
[    7.192526] get_device_id: error, not GC0308
[    7.196719] gc0308: probe of 0-0021 failed with error -1

Probably there is something going on with the I2C GC0308 connection? The I2C connection with the GC0308 goes through two flat cables and connectors: camera and board. Altogether probably 200mm. The shorter the better.

Some user on the badcaps forum reported, that the start-up will work sometimes with low battery.

Maybe it's neccessary to have a look at the signals of that I2C line?

"A low value resistor pulls the SDA and SCL lines strongly to the maximum bus voltage, which could render some bus devices incapable of pulling the line down to the specified low voltage level."

Maybe values of the pull-up resistors overall are too small?

[Fraser]

Slartibartfast123,

Thank you for this valuable input. The issue does seem to revolve around the GC0308 modules communications. I shall shall spend some time investigating the various signals with my DSO. This weekend is pretty busy so it might have to wait until next week. I will try powering the camera from my lab power supply at lower voltages to see if a boot can be completed with a lower Vbat voltage. 

Fraser

[Fraser]

The new GC0308 core arrived today from China. It is a much smaller module, which surprised me. The FPC is not correctly aligned with the camera so the BGA alignment will be off. Basically a junk quality product I did test it with the thermal camera and the original fault persists. Thankfully I do not have to use the Chinese core as we know that problem does not reside in the visible light module itself

Sadly I have been called to a family emergency elsewhere in the country so that will put a delay on the investigation.

Fraser

[Fraser]

Just a quick update on my patient. Nothing found wrong with the I2C communications lines but analysis of the actual data passing along those lines has yet to be completed. More when I have news.

Fraser

[Brexei]

Hi all

I also opened my uti690b and has the same board as in the picture below.
I destroyed my unit by a usb a to micro-usb > usb -c  and/or a quick charge charger. (some how)
I saw on the display 9v 4a! when unplugged it was already done. Not comming on. So be aware off that !
After opening / smelling and seeing with the a flir one cam, the mosfet was gone, next to d5 on the picture of keenox.
It was a 01ah replaced it with a 01A (7 something) and it is working again.
When sensor is disconnected I also get loading 100% and then nothing. after all connected up and running ( so far )
so maybe this can help people.
battery is a pnas 26650 3.6v 5000mah

br
Jeroen

I also have the same problem now!
I switched off the thermal camera, then put it in a different fast charger and.... it was charging...i left it there... then came back and it was making a high-pitch noise. I removed it from the charger and tried to start it... it was dead!!
I am very sad with the fact that there is no info in the manual about this. It seems there is a problem with fast charging.
I opened it up, measured the battery, it is OK. The ON/OFF button is also working fine. But the device does not start at all!
Then i came across your post. I will try to see if the small mosfet is OK.

I am very happy with this camera (despite the small hotspot, that seems to be common to many people) and i really do not want to throw it to the bin.

If anyone else has the same problem please share any info available. I cannot find anything else in the internet.

[Slartibartfast123]

Just a quick update on my patient. Nothing found wrong with the I2C communications lines but analysis of the actual data passing along those lines has yet to be completed. More when I have news.

Fraser

Hm, since someone reported working startup with almost dead battery I would possibly try to power it up with a bench power supply set to 2.7-3.0V - or less.

[Slartibartfast123]

I also have the same problem now!
I switched off the thermal camera, then put it in a different fast charger and.... it was charging...i left it there... then came back and it was making a high-pitch noise. I removed it from the charger and tried to start it... it was dead!!
I am very sad with the fact that there is no info in the manual about this. It seems there is a problem with fast charging.
I opened it up, measured the battery, it is OK. The ON/OFF button is also working fine. But the device does not start at all!
Then i came across your post. I will try to see if the small mosfet is OK.

I am very happy with this camera (despite the small hotspot, that seems to be common to many people) and i really do not want to throw it to the bin.

If anyone else has the same problem please share any info available. I cannot find anything else in the internet.

SMD marking 01AH most likely means IRLML6402TRPbF MOSFET P-Ch -3.7A 65mOhm 8nC Log Lvl.

[Fraser]

Slartibartfast123,

Already tested and sadly no change of symptoms right down to a point where the camera refused to start at all.

For anyone interested, I will be listing the IC’s used in the camera in case a future reader needs the translation of the IC markings etc.

Fraser

[Fraser]

Just a quick update on the poorly 690B camera.

I am carrying out various tests, as and when spare time permits. Today I measured the current draw of the thermal and visible light imaging cores to see if they were within specification.

The thermal imaging core is within specification with a 71mA draw at 3V3 but the the visible light camera is drawing 175mA from the 2.8V supply ! The module should draw around 25mA when running. I measured the resistance across the cores supply rails (whilst disconnected from the camera) and got a reading of 16 Ohms That camera core is most definitely not well.

The question now is why the replacement GC0308 visible light cores did not repair the problem with the camera ? Could they be incompatible in some way ? I do not have experience with these little 24 pin camera modules but I understood them to me interchangeable so long as the connector formats matched. It is possible that some past event damaged the visible light core and that other parts of the 690B camera have also been damaged by the same event. I will do a net list on the 2.8V supply rail that powers the visible light core in case it is power rail related.

I have also noted that the RTC backup power cell is not being charged during camera operation. That is another line of investigation for me.

[Fraser]

In case it is of any interest/use to others, I attach the UART Engineering port 'boot and run' output from the Infiray Tiny1C whilst it is running in the UTi690B. As can be seen, it boots and operates normally, but note that there are communications errors including i2C and failed transfers. It is reporting that a system is "busy" so not communicating with it. We can guess what that means... the host camera controller is not talking with the Tiny1C as it is busy on other matters !

I let the engineering log collection run for a while so that the Tiny1C FFC and self calibration progress may be seen over a period of a few minutes, after a cold start.

For anyone wishing to obtain the engineering log from their UTi690B/290B you will find the UART port annotated (GND,TX,RX) on the thermal/visible light core PCB. The serial settings are 57600 Bd, 8,N,1 no handshake and 3V3 levels.

Fraser

[Fraser]

The more I look at the behaviour of the poorly UTi690B, the more I think that the operating system is getting its 'knickers in a twist' when the GC0308 visible light camera core does not respond as expected. I have proven that the original GC0308 camera is faulty and it even gets quite warm due to the internal low impedance across its supply rails. In such a situation you might expect that the UTi690B camera just needs a replacement GC0308 core, and all will be well again. So often life is not that simple though and this appears to be such a case. It could just be that the the two versions of GC0308 core that I bought are not identical to the original fitment and so not acceptable to the operating system. I can find no fault on the visible light core 2v8 power rail. That power rail is dedicated to just the visible light core. It is provided by a DC-DC buck converter U3 on the sensor core PCB from the 3V3 power rail.

It would be unusual to have a faulty visible light core plus an additional fault in the systems communications bearers (I2C) but I never say never.

At this stage of the investigation I have been unable to find any other significant faults with the UTi690B camera. I am investigating the RTC backup power supply but it may well be that I am expecting more voltage across the cell than is actually needed (I have 2V9 across it)

My next move will be analysis of the GC0308 core situation..... communications signals and a search for an identical model of GC0308 based camera core.

Fraser

[Fraser]

UTi690B/290B supply rails and their sources

Uni-T helpfully provide annotated test points for all of these voltages on their PCB's.

Main PCB

3V3 from U6 63025P
1V2 from U7 PD5Q
5V0 from U9 SF9B
1V35 from U8 PD5Q

Imaging cores PCB

1V8 from U2 PD5Q
2V8 from U3 NL

[Fraser]

The UTi690B/290B Chipset identities

D9SHD - MT41K256M16TW-107 - 4Gb DDR3 SRAM
MCIMX6Y0CVM08AB - NXP i.MX-6ULL Application Processor with ARM CORTEX A7
NL - TLV700XX - 2V8 200mA LDO Regulator
P8563 - PCF8563 - Real Time Clock
PD5Q (U7) - TPS62087 - 1V2 3A step down converter
PD5Q (U8) - TPS62087 - 1.35V 3A step down converter
PD5Q (U2) - TPS62087 - 1v8 3A step down converter
SF9B - LMR64010 - 5V0 1A step up simple switcher voltage regulator
AVC16245 - SN74AVC16245 - 16bit tri-state bus transceiver
THGBMNG5D1LBAIL - 4GB e-MMC flash memory module
TP5000 - 2A switching single Lithium Cell Charger
63025P - TPS630250 - 3V3 2A variable output buck-boost converter
V5L -  Not identified at this time. There are several of these on the PCB so likely provides simple functionality

UPDATE: V5L looks to be a 5V rail (6V breakdown) Transient Voltage Suppressor, namely the "Way On" WE05-4RVLC

[Ed.Kloonk]

How does one examine a IR camera board for heat spots when it's the camera itself under investigation? Asking for a friend.

[TobyG]

lol, the old fashioned way, I guess? IPA and see where it evaporates 

[Slartibartfast123]

Terrific job so far on the poor thing, Fraser. 

The thermal imaging core is within specification with a 71mA draw at 3V3 but the the visible light camera is drawing 175mA from the 2.8V supply ! The module should draw around 25mA when running. I measured the resistance across the cores supply rails (whilst disconnected from the camera) and got a reading of 16 Ohms That camera core is most definitely not well.

Did I get this right? There is a 16 Ohms short on that specific power rail while the visible light camera is disconnected? Grab another thermal camera and have a look under power. Maybe there is a shorted ceramic cap – not uncommon these days.

[Fraser]

No, there is a low impedance on the actual GC0308 visible light camera module power input pins. With the GC0308 removed from the camera, it measures only 16 Ohms across its power input pins and draws 175mA when connected to 2.8V. With the GC0308 removed from the camera there is no issue with current draw on the host UTi690B 2.8V power rail. The low impedance is definitely in the GC0308 module and it gets quite warm when powered. The GC0308 should draw only 25mA when operating normally.

[Fraser]

Further investigation of the faulty GC0308 core module will follow. I wonder if it has a decoupling capacitor across its supply rails hidden from view in the modules casing ? A failed MLCC certainly fits the symptoms as has been said, but it is not on the PCB so it makes sense to carefully dissect the GC0308 core.

Fraser

[Fraser]

The result of the GC0308 core dissection.....

The GC0308 visible light imaging core does indeed contain an MLCC power rail decoupling capacitor !

Upon testing the suspect GC0308 I found the resistance between the 2V8 and 0V power rails to be 16 Ohms. Once the suspect MLCC was removed from the FPC, the low impedance was also removed from across the cores 2v8 power rail so the GC0308 core should be operational 

The bad news is that the camera will still not boot with this GC0308 module fitted 

For other owners of this camera who have issues with it not booting, it would be worth checking the GC0308 power supply rails as this MLCC failure may be quite common ? I attach a picture of the GC0308 core after the MLCC was removed. It was located below the FPA in this image and the two solder pads are visible. It is a tiny component.

As to why my patient still refuses to complete boot despite removal of the faulty component....I am unsure why this is, but it may explain why the replacement GC0308 core did not solve the boot problem. Something else has failed on this camera. 

Fraser

[Slartibartfast123]

Interesting news! Hm, several reasons for not functioning:

1. When the camera module went bad it also somehow damaged the 2V8 power rail of the board. Not necessarily the case.

2. When the camera module went bad it damaged the power rail and/or somehow shorted the I2C lines and damaged something.

3. The camera module needs a working decoupling capacitor for proper functionality. There is a reason why they put it so close to the FPA. Most likely it won't work good without any cap.

4. The original camera module is dead by now (short & overheating) and other camera module is not compatible for some reason.

My bet would be on 3 & 4.

The datasheet of the camera module most likely looks like this (pin 1, 23, 24: NC, pin 2: GND, pin 4: VDD 2.8V):