Troubleshooting Circuitry in Old SD49225T-HN PTZ Cameras

[EDIT] Will refer to this camera as Alpha for this thread [End EDIT]

Reporting a successful fix of an ailing Dahua SD49225T-HN. It wasn't suffering the dreaded rebooting, but was beginning to routinely lose IP communications for a few seconds after any large, rapid PTZ maneuvers. Basically, it was acting as though the PTZ motors were dragging down the power supply enough to disrupt IP communications. It would always recover, but rapidly becoming unusable as a PTZ. After all, what good is a PTZ camera if it cuts out for 10-20 seconds after you change its direction?

I took off the top dome from camera to examine the POE power board. I could see one electrolytic cap on its top surfaces. That was a 63 volt 50 uFd unit. It measured (in circuit) at 47 uFd.



On the bottom surface of the POE board, I found a small sized 16 volt, 100 uFd capacitor. That measured only 80 uFd, a limits of acceptable tolerance. Also, there was a jumper wire that had been placed at the factory. Surmising that this little 100 uFd capacitor was probably one for filtering the POE board output, I decided to try replacing it.




I didn't have an exact match in my drawer, but happened to have a good supply of 16 volt 1000 uFd caps. Typically, one doesn't make that big of a capacitance change, but nothing to lose....
Here is the new capacitor soldering in place of the old one. I made sure to observe proper polarity on the electrolytic cap and insulate its leads.



Put it all back together and reinstalled the camera. Repeated PTZ moves to greatly differing preset positions work fine. Manual maneuvering also without inducing the aforementioned loss of IP communications.
This may have solved it. It certainly is a lot better than it was before the capacitor replacement.

Interestingly, the POE power board on another camera did not have the jumper wire and capacitor on its underside. Were the jumper and capacitor a hot fix for a undersized POE power supply design? It may well be so.

May have spoken a little too soon. Indeed, there are no longer IP dropouts, but the IR illuminator sometimes kicks off and then back on with some PTZ movements. It is definitely better, but there is still something not quite right.

[EDIT] Correction. It is not an IR dropout, but rather an effect of using a shutter range causing dimming of image if camera encounters a big change in scene brightness [End EDIT]

It is not enough to have their value, they have an ESR value as well, and you need an ESR meter to check that. That value could be much higher than a new one. Did you replace the 47uf one? If not, maybe you should.

tibimakai said:

you need an ESR meter to check that

Click to expand...

I agree. Unfortunately I could not get a clean read of ESR, in circuit, with my DE-5000. DIdn't really matter. That PTZ still has issues even after transplatnting another POE board from a well functioning unit.
Will have to dig further into the camera module boards, but that will have to be another day.

You don't have a thermal gun by any chance? Maybe that would show something overheating.

I do have a thermal camera, even added a ZnSE lens to enable macro focusing for circuit boards.
My 3D print design here Macro Lens and Flip Up Cap for HTI-19/HTi-18 Thermal Imaging Camera by BunnyScience

(Thank you Steve, Diana, & Tia - never forgotten)

Digging deeper into another of my problematic SD49225T-HN's. This one mainly had reboot problems and would not store presets past reboots. I could store new presets and they would work UNTIL the next reboot. Then it would always revert to some old saved presets, instead of the ones it had dutifully accepted.

Took things further apart to look at the stepper motor board. That is accessed by removing the side covers and four screws. That frees the main camera body from the dome and exposes the stepper drive board (at least that is what I think it is). That board had two 16 volt electrolytics. One 100 mfd, the other 470 mfd. Only the 470 mfd had slightly high ESR at 0.25, but I went ahead and replaced both of them while I had that board out of camera. Those probably buffer the power while the stepper motors do their thing. I'll probably go back to my partially fixed one and try replacing these two caps to see if the problem with IR lights post PTZ motions gets fixed.

The most interesting finding was when I looked at the camera module of this amnesiac camera.

This model camera may have a tiny, soldered on memory battery. If so, that battery going dead would explain failure to properly store presets.


Seen up close on my bench microscope it has polarity markings. My DVM read zero volts across its terminals. So it is either dead or not a battery.
M82D1 doesn't come up with any battery on google.


I guess I could try measuring the voltage across the potential "battery" on a camera that still allows storage of settings and see if it is greater than zero.

If you're interested, here's some information from me: As I have already written in my other postings, I have now connected my old SD49225T-HN (Total Working Time: 2223 days) in my study for testing. There it is connected with a 3A power supply and not via PoE. Since I have the camera here in the study, the reboots have definitely become fewer.



I can't say whether this is because the camera is connected to a power supply and not via PoE, or perhaps because there is no humidity here in the study.

Since I have integrated my Dahua cameras into Home Assistant via Onvif, I noticed that the processor utilization on the SD49225T-HN is 100% more often and over a longer period of time. This is not the case with the SD49225XA-HNR, for example. Both cameras are set up virtually identically and detection is only carried out via IVS Line Detection.






I can't judge whether the 100% CPU utilization on the SD49225T-HN means anything, but that shouldn't be normal either.

I already had the problem that the PTZ function of the SD49225T-HN went crazy and the camera rotated constantly, but this has only happened very rarely so far. Maybe once every four or five months.

Since there are also reboots for the power supply via power supply - although fewer now - I actually rule out a problem with PoE. The fact that there are now fewer reboots here in the study could have something to do with the lack of moisture. But that's just speculation on my part. I will keep the camera running here in the study until the end of the month and continue to monitor the topic. I will then mount the camera outside again as a test and then see if the reboots occur more frequently again.

Actually, the camera is still too good to throw away.

That's good additional behavioral information. It does suggest the POE board deteriorating over time - which is what I expect will happen. Your bypassing the POE board using a 12 volt supply and seeing fewer reboots probably means the POE board supplies lower and lower voltage as it ages. That then pushes components further down the chain to fail. Those too are aging and less able to tolerate reduced supply.

When thinking of age related failures, the usual suspects are...

1. Electrolytic capacitors - these have a finite lifespan. In extreme temperature environments like our PTZ cameras, a 5 to 7 year lifespan would not be surprising.
2. Embedded 3 volt batteries. Some devices have these and behave very oddly once the battery dies. These are typically used to keep data alive in semi-permanent storage.
3. Non-volatile memory exceeding write endurance.
4. Regulation diodes, power resistors going out of spec

Of the above, the first two are the easiest to find and attempt replacement.

There is a limit to how much effort I'm willing to put into fixing these up. With newer, more capable replacements available, going into super detail and checking every component isn't worth my time. A couple pokes for possibly fixing and my curiosity is where I am at with these old PTZ's.

Pulled entire camera module out to get better access. It's only three cables and four screws, but the PITA is the double sideded taped in place silica gel packs. One ruptured during removal. Well, need to replace them anyways before re-sealing camera.



After desoldering the tiny button cell and tearing off its welded on tab, I can read some markings on the cell. The ruler markings you see next to the cell are mm. This is a tiny button cell of some sort.



It is a Seiko ML Lithium Rechargeable Battery ML414H

SPECIFICATIONS
Nominal Voltage(V) 　3
Charge Voltage*1 (V)
(Standard Charge Voltage) 2.8 to 3.1
Nominal Capacity
(Voltage Range V)(mAh) 1.0 (3.1V ～2.0V)
Internal Impedance*2 (Ω) 600
Standard Discharge Current(mA) 0.005
Cycle Life *3 (Time)50%
D.O.D(Depth of Discharge) 300 (10% D.O.D*4)
Diameter(mm) 4.8
Height(mm) 1.4
Weight(g) 0.07

Bolded for emphasis. This is a rechargeable 3 volt cell. We cannot replace this with a standard lithium cell. That would possibly explode if camera attempts to charge it.
Notice how short the cycle life is on the cell. ONLY 3 half discharge cycles!!!!

Mouser has these for about $2 each. Might be worth trying to replace on a couple of my sick PTZ's at that price. I'll bet shipping will be more than the parts.

Ordered up some to test in my sick PTZ's (the ones that lose their PTZ settings between bootups. Given that this type of cell is commonly used to maintain SRAM, and the cell I pulled is completely dead, it is a reasonable thing to try for that particular symptom.

The POE power degradation issues are a separate matter that would also need to be addressed. Beyond replacing the caps on the POE board and PTZ board and also this rechargeable button cell, I doubt it's worth delving deeper.

BTW removing the button cell will erase everything that is in SRAM. If Dahua does not also have a ROM in the design, this will also mean needing to write firmware back into the camera. I have not ever done that process.
At any rate, the cell was already dead. So, I don't know exactly how high the risk actually is of the firmware being totally gone.

[EDIT] Removing the battery does not brick the camera. Firmware seems to survive intact even with the battery removed for a week. Camera won't function without the battery, but it comes back to life when new battery was installed. [End EDIT]

I wonder if a super cap would last longer

A really small super cap

Since this really sick PTZ [Bravo] is awaiting a new cell battery, I transplanted its recapped PTZ driver board into my camera Alpha (the one first mentioned in this thread) that has troubles post PTZ moves. Recall I recapped its POE board and got some improvement, but it was dropping out its IR illuminator post PTZ moves. We'll see if the recapped PTZ driver board helps that issue tonight.

It was convenient to do the transplant now because I needed to work on the wiring there to add two new turrets atop that mounting pole.

[EDIT as it turns out Alpha was actually working correctly. I was misinterpreting it dimming the image after PTZ moves as IR dropout. The first POE electrolytic capacitor swap had already fixed it signal dropout problem [End EDIT]

I also thought of something else: With my SD49225T-HN there is also the effect that the IR switching (day/night) sometimes doesn't work. But that also happens rarely, at least much less often than the restart.

No IR switching = approx. once every two months
Automatic restart = several times per week
PTZ restart = approx. 1 to 2 x per month
Camera has hung up and needs to be disconnected from power = approx. once every two months

In any case, I'm curious whether changing the CMOS battery will help.

That little "battery" looks almost like a super cap, that I have just replaced in my Fluke 189.
With the thermal gun, you did not notice anything red hot? Like an SMD capacitor?
I have found in the past shorted SMD caps, causing the IR to not work.
Greta job, of working on these.
When I have got mine a few years ago, the IR did not work, I took it apart, and it had the IR light connector a bit loose. I did all this with Andy's approval.

The physical layout isn't conducive to viewing the boards in operation. Camera boards are pretty well encased when system is put together enough to power up.
How did you accomplish looking at the camera boards in operation?

I have never repaired one of these, just basic cameras.

Hello everyone, it seems that I have come to the right place so that someone can shed some light on my problem. I have a Dahua SD49225T-HN-S2 PTZ camera that stopped working overnight, entered a reboot loop and returned to the factory IP 192.168.1.108 Unfortunately I cannot fix it using TFTP. It keeps rebooting and the connection cannot be established. I want to find the UART port of the camera but I can't find it. I only found these single pins that appear to be the UART port, but when connecting them with a USB-TTL nothing appears on the monitor when power is connected to the camera.

• I have always used a 12V 3A power supply and not a POE supply
• I have measured the SRAM battery on the circuit-board, I have not removed it yet but it seems dead, it measures 207 mV
• When connecting the power source to the camera, the Ethernet LEDs work after 5 seconds but after another 10 seconds they turn off and restart in a loop. (When I ping the IP address 192.168.1.108 I have a response but then it disappears)

Jim_OS said:

Automatic restart = several times per week

Click to expand...

I'll now give up looking for the cause of the reboot error. There are days, even two or three days in a row, when the error doesn't occur and then there are days, like yesterday, when the error happens three or four times.


On February 12th and 13th there were no reboots. On the days before, however, it did.

I can now rule out a moisture problem and it's also not due to the PoE power because I'm currently using the camera with a 12V 3A power adapter.

I could try the firmware from 2019 again, but unfortunately there is an (API) problem in connection with Home Assistant. I don't use motion detection and therefore it is deactivated in the camera's WebGUI. Every time Home Assistant reboots, motion detection in the camera's WebGUI is automatically reactivated. This happens both with the Dahua addon from rroller, but also if the camera was integrated into Home Assistant via Onvif integration. Why that is: I have no idea. In any case, I can't use the old firmware from 2019 either.

Maybe and hopefully guykuo or another user will find the exact reason for the errors, but as I said, I'll give up further searching for now.