OK some comment on the TAU core crash that is detailed in this thread.
The core had crashed and locked up for reasons unknown. It would not activate its FFC shutter at boot and it was not producing a video image. It was effectively in a HALT condition.
Through fault exploration of the TAU circuit I was able to be confident that all supply rails and chips on the PCB were in a state where they should be working. This just left a firmware corruption or other similar software related cause for the cameras failure to produce an image. The firmware is not easily accessible on the TAU but if the firmware is intact and just a camera reset is needed, there is hope. I got some help with resetting the camera to a point that it would talk over its RS232 port. Once I was able to talk to the camera via its GUI, I was able to see that the FFC shutter was set to"External". That is to say, a external controller was telling the TAU when to generate an FFC event. Remember that this HS324 can record video so I an guessing that the external ARM processor that drives the TAU controls the FFC event so that it can stop the FFC whilst in video recording mode.
Why did the camera crash ? At first I blamed a known problem with early firmware but having re-assembled the HS324 to the point that i could test the complete system, I now believe there is another issue at work here.
I have set the TAU to internal auto FFC as a way of monitoring its boot sequence, and making it independent of the rest of the HS324. The HS324 was switched on and the normal boot screen of a colour bar chart appeared. The TAU booted and carried out the normal two FFC cycles as it should when set to AUTO FFC. All was looking promising. The HS324 has three power modes, OFF, Standby and ON. From Standby to ON takes but a second or two for an image to appear. From OFF to ON takes 90 seconds. I waited the 90 seconds and heard some odd behaviour from the TAU. It was making a varying whine and the FFC shutter sequence was repeating. I striped the chassis off of the TAU and the PCB's that attach to its rear so that voltage checks could be carried out. I could immediately see that the TAU was in a ON/RESET loop as the 3V3 rail monitor RESET LED and FPGA boot LED were blinking alternately. No wonder the TAU had crashed if it was in a ON/RESET loop for very long. Such is a recipe for disaster.
With the TAU power input rail now accessible I could see that the 5V rail was all over the place with very poor stability and dropping as low as 3V. The poor 3V3 power rail monitor IC had no choice but to reset the camera every time the supply dipped below the 3V3 buck converters minimum operating voltage. It is the buck converters that are producing the noise as they try to maintain their power rails. The sound is barely perceptible unless you place your ear right next to the PCB.
It is now pretty obvious that the poor TAU was a victim of a highly unstable 5V supply rail. I suspect that same unstable supply rail is causing similar issues for the ARM chip that controls the TAU. Hence no boot completion. Its supply monitor will also be having a minor fit if the supply keeps dropping. I shall investigate the cause of this irregular supply rail but at least I have something to chase down and a possible, ney likely, cause of the TAU crash.
Sadly I did not previously monitor the supply rails on the HS324 and the add on modules that attach to the rear of the TAU as my priority was to get the TAU functioning again. Knowing that the supply rail to the TAU was compromised would have helped explain the crash and I would not have searched so hard for a hardware fault in it that was not present. You live and learn eh
I did learn a lot about the TAU hardware in the process though and I enjoyed exploring it so no worries really. The camera did need to be recovered from its crashed state anyway.
The next steps will be to check the power main power supply rail into the cameras 'Power & Video PCB' and trace its through the various Boost/Buck converters that are present in the modules. The fault could be a component failure that is leading to an overload of the 5V supply rail.
I did a quick check across the 4V8 Ni-Mh pack and was surprised to see it alternating between 4V94 and 3V6. Something nasty is going on. The rail cycles in sync with the TAU trying to Boot and I know the TAU current draw is correct. Something may be causing an over current event soon after camera power is applied. It is just a case of tracking down the guilty component
It could, of course be a poor connection in the sealed battery bay so that will be checked fist by replacing it with a current limited bench power supply.
All good fun.
Fraser
Home
Help
Search
Login
Register
