Typical Dahua cameras failures?

[tibimakai]

There is a coin cell battery on the camera board, 2.7V. Appears to be spot welded to leads and soldered to board. Might be the 3V nominal version of the below, with 100 cycle life.

https://www.mouser.com/new/seiko-instruments/seiko-ms-lithium-battery/

I would test for shorted smd capacitors. I have already fixed two other brand cameras, by just removing the shorted smd capacitors. Usually they are the bigger ones, that filter the power line. There are multiple in parallel.

 

[carbonita]

@tibimakai Thanks -- will do next, for science! FYI, PoE board image

 

[tibimakai]

The bad smds were on the main board. Both cameras were Swann cameras.

 

[carbonita]

It's Alive! Tested all the caps that I could get probes on (some SMD parts are tiny), and all were good. Turns out the cam spontaneously bricked, as no changes were done to the system during the time it intermittently failed. Used @cor35vet instructions from "Dahua IPC unbricking / recovery over serial UART and TFTP" thread to revive, and used older fw "13", vs the "15" used at time of self-bricking. Note that fw version "16" failed to revive cam. Hopefully stays stable and operational. Image below may be useful for others to locate TX/RX holes.

 

[tibimakai]

The 16 bricked my wedge cam, but I have managed to revive it, and I have the 16 on it right now and it works as my other Dahua cameras.

 

[carbonita]

New failure mode? The recording interval spontaneously (untouched camera) to a fixed 8 minutes until reboot. This fixed interval recording starts from 10-12 hours after reboot, so daily reboots are insufficient. BTW, this camera has always required daily reboots since purchase in Oct2018 to reset the IR cut filter. Note the change in event-activated recording to fixed interval recording in the attached screenshot. The interval is the Storage > RecordControl > PackDuration from my fiddling with this parameter, and changing from auto to manual does not remedy.

 

[carbonita]

Furthering the mystery, using VBR, or smartCodec, or H265, or bitrates higher than 1280kbps appears to correlate to failure modes. Possibly thermal/overheating problem, and I wonder if the oozing thermal interface pads are gapped from the aluminum housing heatsink.

 

[carbonita]

Its a thermal runaway problem. Using thermal paste, interface pads, and aluminum sheets resolved the issue for daytime setting, covering only the Ambarella SoC and adjacent IC. However, for nighttime setting, the IR circuit components generate more heat, which is resolved by running a fan while the camera boards are laid bare on the bench. Hoping to resolve with better thermal interface pad to sink the entire board to the support and case. BTW, if anyone wants the SMD part numbers, I'll provide.

 

[TheDude]

One of my SD1A203T PTZ cameras will only work when powered via the 12V power jack now. Will not power up with POE using the NVR5216-4KS2 NVR or with a Texas POE injector.....

 

[CCTVCam]

It's Alive! Tested all the caps that I could get probes on (some SMD parts are tiny), and all were good. Turns out the cam spontaneously bricked, as no changes were done to the system during the time it intermittently failed. Used @cor35vet instructions from "Dahua IPC unbricking / recovery over serial UART and TFTP" thread to revive, and used older fw "13", vs the "15" used at time of self-bricking. Note that fw version "16" failed to revive cam. Hopefully stays stable and operational. Image below may be useful for others to locate TX/RX holes.

I presume you using VPN as sounds more likely it was hacked.

 

[carbonita]

Air gapped. Are Dahua's particularly susceptible to hacking?

I presume you using VPN as sounds more likely it was hacked.

 

[carbonita]

The PoE buck converter is a separate board in my camera. Maybe the same on your cam. There are two electrolytic caps, which seem to have low MTTF, the remainder being SMD. Straightforward to do a continuity check.

One of my SD1A203T PTZ cameras will only work when powered via the 12V power jack now. Will not power up with POE using the NVR5216-4KS2 NVR or with a Texas POE injector.....

 

[tibimakai]

Most likely one of those smd ones is shorted. Just measure them with a multimeter(w/beep).

 

[carbonita]

For other DIYers that may want to save a few dozen bucks without the hours of debugging entertainment , the cam has been working fine with SmartIR off for a week, and manually set at 50% IR at night. With external IR illuminator, plates are captured as well as posted in LPR forum a couple of years back. To try to get SmartIR back, or even 100% IR, here's the addition of (blue) thermal pad, paste, Aluminum sandwich over the IR smd circuitry.