System Red Amplifier probs

[ShimmerNZ]

I've been having problems with my red amplifier recently. Symptoms are working fine then all of a sudden lights on controller dim and power drops on powered usb hub. Same symptoms when powered from PC usb port and similar behavior when powered off a battery.

When unplugging the amplifier it comes back to life. I can see similar behavior if I unplug from controller and power via mini usb port so reasonably sure its the amp. I have used a different power cable and LAN cable as part of troubleshooting

I've checked all solder joints and they are all good. There was no water ingress in the box it was mounted in. I thought it might be have been the electrolytic capacitors so replaced all 4 of them but still not solved.

I've noticed on the webpage it comes up with:
No or unknown amplifier connected

In the amplifier section it shows

PCB 12.3 FW 1.7

and I can see it disappear , reappear then drop the gain in steps as adc voltage too high, settled and then disappear.

At a loss what to try next.


log looks like this...

Code:

Blitzortung.org Controller
Station 1243
FW 8.2PCB 10.4/12.3
2017-04-17 06:57:14   9138 | AUTO 1/1: [NOISE IS LOWER] -32/30mV. Trying to raise gain...
2017-04-17 06:57:14   9138 | AMP1: Set gain 6 on channel 0, op-amp 0
2017-04-17 06:57:14   9138 | AMP1: Set gain 6 on channel 0, op-amp 1
2017-04-17 06:57:14   9138 | AUTO 1/2: [NOISE IS LOWER] -20/19mV. Trying to raise gain...
2017-04-17 06:57:14   9138 | AMP1: Set gain 6 on channel 1, op-amp 0
2017-04-17 06:57:14   9138 | AMP1: Set gain 6 on channel 1, op-amp 1
2017-04-17 06:57:14   9138 | AMP1: Ch1, Gain1:     3   ->  6
2017-04-17 06:57:14   9138 | AMP1: Ch1, Gain2:     0   ->  6
2017-04-17 06:57:14   9138 | AMP1: Ch2, Gain1:     3   ->  6
2017-04-17 06:57:14   9138 | AMP1: Ch2, Gain2:     0   ->  6
2017-04-17 06:57:14   9138 | AMP1: Set gain 5 on channel 0, op-amp 0
2017-04-17 06:57:14   9138 | AMP1: Set gain 5 on channel 0, op-amp 1
2017-04-17 06:57:14   9138 | AMP1: Set gain 5 on channel 1, op-amp 1
2017-04-17 06:57:14   9138 | AMP1: Ch1, Gain1:     6   ->  5
2017-04-17 06:57:14   9138 | AMP1: Ch1, Gain2:     6   ->  5
2017-04-17 06:57:14   9138 | AMP1: Ch2, Gain2:     6   ->  5
2017-04-17 06:57:19   9143 | AMP1: Command 'Board 1, Firmware 1.7'
2017-04-17 06:57:19   9143 | AMP1: Controller-board 'Nice to meet you. :-)'
2017-04-17 06:57:20   9144 | AMP1: Setting all gains...
2017-04-17 06:57:21   9145 | EVENT: Amplifier 1 is now connected
2017-04-17 06:57:21   9145 | === New Amplifier 1 Connected ===
2017-04-17 06:57:21   9145 | AMP1: PCB:           -1   ->  1
2017-04-17 06:57:21   9145 | AMP1: Firmware:       0.0 ->  1.7
2017-04-17 06:57:21   9145 | AMP1: Refresh-Intvl:  0   ->  0
2017-04-17 06:57:21   9145 | AMP1: Localmode:      0   ->  1
2017-04-17 06:57:21   9145 | AMP1: Channels:       0   ->  2
2017-04-17 06:57:21   9145 | AMP1: Ch1 op-amps:    0   ->  2
2017-04-17 06:57:21   9145 | AMP1: Ch1, Gain1:     5   ->  0
2017-04-17 06:57:21   9145 | AMP1: Ch1, Gain2:     0   ->  0
2017-04-17 06:57:21   9145 | AMP1: Ch2 op-amps:    0   ->  2
2017-04-17 06:57:21   9145 | AMP1: Ch2, Gain1:     5   ->  0
2017-04-17 06:57:21   9145 | AMP1: Ch2, Gain2:     0   ->  0
2017-04-17 06:57:22   9146 | AMP1: Setting all gains...
2017-04-17 06:57:22   9146 | AMP1: Set gain 5 on channel 0, op-amp 0
2017-04-17 06:57:22   9146 | AMP1: Set gain 0 on channel 0, op-amp 1
2017-04-17 06:57:22   9146 | AMP1: Set gain 5 on channel 1, op-amp 0
2017-04-17 06:57:22   9146 | AMP1: Set gain 0 on channel 1, op-amp 1
2017-04-17 06:57:22   9146 | AMP1: Localmode:      1   ->  0
2017-04-17 06:57:22   9146 | AMP1: Ch1, Gain1:     0   ->  5
2017-04-17 06:57:22   9146 | AMP1: Ch2, Gain1:     0   ->  5
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Message with status 32, seconds 120, text 'Error: Invalid amplifier PCB detected! '
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Message with status 32, seconds 120, text 'Error: Invalid amplifier PCB detected! '
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Message with status 32, seconds 120, text 'Error: Invalid amplifier PCB detected! '
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Message with status 32, seconds 120, text 'No or unknown amplifier connected! '
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Server '0' deactivated
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Server '0', expert '0'
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Set max relative gain: rgain_max=10240, threshold_min=100 on amp0, ch0
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Set max relative gain: rgain_max=10240, threshold_min=100 on amp0, ch1
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Set max relative gain: rgain_max=0, threshold_min=100 on amp0, ch2
2017-04-17 06:57:23   9147 | REMOTE-CONFIG: Done with 10 updates and 0 errors
2017-04-17 06:57:24   9148 | AUTO 1/1: [NOISE MUCH TOO HIGH] 12692/12762mV. ADC accuracy could be affected! Trying to lower gain...
2017-04-17 06:57:24   9148 | AMP1: Set gain 3 on channel 0, op-amp 0
2017-04-17 06:57:24   9148 | AUTO 1/2: [NOISE MUCH TOO HIGH] 12706/12743mV. ADC accuracy could be affected! Trying to lower gain...
2017-04-17 06:57:24   9148 | AMP1: Set gain 3 on channel 1, op-amp 0
2017-04-17 06:57:24   9148 | AMP1: Ch1, Gain1:     5   ->  3
2017-04-17 06:57:24   9148 | AMP1: Ch2, Gain1:     5   ->  3
2017-04-17 06:57:30   9154 | REMOTE-CONFIG: Server '0' activated
2017-04-17 06:57:30   9154 | REMOTE-CONFIG: Server '0', expert '1'
2017-04-17 06:57:30   9154 | REMOTE-CONFIG: Done with 3 updates and 0 errors
2017-04-17 06:57:34   9158 | AUTO 1/1: [NOISE IS LOWER] -32/31mV. Trying to raise gain...
2017-04-17 06:57:34   9158 | AMP1: Set gain 6 on channel 0, op-amp 0
2017-04-17 06:57:34   9158 | AMP1: Set gain 6 on channel 0, op-amp 1
2017-04-17 06:57:34   9158 | AUTO 1/2: [NOISE IS LOWER] -19/20mV. Trying to raise gain...
2017-04-17 06:57:34   9158 | AMP1: Set gain 6 on channel 1, op-amp 0
2017-04-17 06:57:34   9158 | AMP1: Set gain 6 on channel 1, op-amp 1
2017-04-17 06:57:34   9158 | AMP1: Ch1, Gain1:     3   ->  6
2017-04-17 06:57:34   9158 | AMP1: Ch1, Gain2:     0   ->  6
2017-04-17 06:57:34   9158 | AMP1: Ch2, Gain1:     3   ->  6
2017-04-17 06:57:34   9158 | AMP1: Ch2, Gain2:     0   ->  6
2017-04-17 06:57:34   9158 | AMP1: Set gain 5 on channel 0, op-amp 0
2017-04-17 06:57:34   9158 | AMP1: Set gain 5 on channel 0, op-amp 1
2017-04-17 06:57:34   9158 | AMP1: Set gain 5 on channel 1, op-amp 1
2017-04-17 06:57:34   9158 | AMP1: Ch1, Gain1:     6   ->  5
2017-04-17 06:57:34   9158 | AMP1: Ch1, Gain2:     6   ->  5
2017-04-17 06:57:34   9158 | AMP1: Ch2, Gain2:     6   ->  5
2017-04-17 06:57:37   9161 | AMP1: Command 'Board 1, Firmware 1.7'
2017-04-17 06:57:37   9161 | AMP1: Controller-board 'Nice to meet you. :-)'
2017-04-17 06:57:38   9162 | AMP1: Setting all gains...
2017-04-17 06:57:38   9162 | EVENT: Amplifier 1 is now connected
2017-04-17 06:57:38   9162 | === New Amplifier 1 Connected ===
2017-04-17 06:57:38   9162 | AMP1: PCB:           -1   ->  1
2017-04-17 06:57:38   9162 | AMP1: Firmware:       0.0 ->  1.7
2017-04-17 06:57:38   9162 | AMP1: Refresh-Intvl:  0   ->  0
2017-04-17 06:57:38   9162 | AMP1: Localmode:      0   ->  1
2017-04-17 06:57:38   9162 | AMP1: Channels:       0   ->  2
2017-04-17 06:57:38   9162 | AMP1: Ch1 op-amps:    0   ->  2
2017-04-17 06:57:38   9162 | AMP1: Ch1, Gain1:     5   ->  0
2017-04-17 06:57:38   9162 | AMP1: Ch1, Gain2:     0   ->  0
2017-04-17 06:57:38   9162 | AMP1: Ch2 op-amps:    0   ->  2
2017-04-17 06:57:38   9162 | AMP1: Ch2, Gain1:     5   ->  0
2017-04-17 06:57:38   9162 | AMP1: Ch2, Gain2:     0   ->  0
2017-04-17 06:57:39   9163 | AMP1: Setting all gains...

[BobW]

I think you've definitely pinned it down to the AMP. Something appears to be dragging the +5 down. Trick is finding what, and there really isn't too much that can do it. There's like 8 chips and the micro. I think if the micro was doing it, it wouldn't recover. So, i'd power it from something that could deliver some current, but not enough to burn traces. Maybe 4 AA batteries through a diode -1n4001 or something - to drop the 6 volts to 5.3 and a 10 ohm resistor to keep the current down to less than a 1/2 amp if hard short. Your USB supplies - charger or PC - are going in to hard limit, and not putting much current out if they see a short. 

So, something dragging the +5 down, if you can let it draw some current, will make heat, and be easy to find. Be careful when using your finger to find it so as not to burn the chip logo into the blister. With the powering arrangement I described, it might become self evident via it getting hot enough to smoke. If so, make sure you don't let it be hot long enough to burn the board. The amp normally draws nearly nothing, so it working properly, it should run fine powering it through the arrangement below. If something drags the 5 volt line down, that thing should get hot, but 1/2 amp shouldn't be enough to damage traces. You could also probe around with a meter to the different parts on the board that get +5 and find the minimum point referenced to the - power input. Do not just connect it to a high current power source - nicad or power supply - without the resistor, as you will smoke a trace if the short is hard enough on the board. 


                                                               Diode    10ohm
                                                       + ------|>|----/\/\/\/------- + to amplifier
   (Battery or other 5-6 volt source) - ----------------------------- -


You probably have it, but schematic on page 11
http://en.blitzortung.org/Compendium/Doc...B_12.3.pdf

[ShimmerNZ]

I think you've definitely pinned it down to the AMP. Something appears to be dragging the +5 down. Trick is finding what, and there really isn't too much that can do it. There's like 8 chips and the micro. I think if the micro was doing it, it wouldn't recover. So, i'd power it from something that could deliver some current, but not enough to burn traces. Maybe 4 AA batteries through a diode -1n4001 or something - to drop the 6 volts to 5.3 and a 10 ohm resistor to keep the current down to less than a 1/2 amp if hard short. Your USB supplies - charger or PC - are going in to hard limit, and not putting much current out if they see a short. 

So, something dragging the +5 down, if you can let it draw some current, will make heat, and be easy to find. Be careful when using your finger to find it so as not to burn the chip logo into the blister. With the powering arrangement I described, it might become self evident via it getting hot enough to smoke. If so, make sure you don't let it be hot long enough to burn the board. The amp normally draws nearly nothing, so it working properly, it should run fine powering it through the arrangement below. If something drags the 5 volt line down, that thing should get hot, but 1/2 amp shouldn't be enough to damage traces. You could also probe around with a meter to the different parts on the board that get +5 and find the minimum point referenced to the - power input. Do not just connect it to a high current power source - nicad or power supply - without the resistor, as you will smoke a trace if the short is hard enough on the board. 


                                                               Diode    10ohm
                                                       + ------|>|----/\/\/\/------- + to amplifier
   (Battery or other 5-6 volt source) - ----------------------------- -


You probably have it, but schematic on page 11
http://en.blitzortung.org/Compendium/Doc...B_12.3.pdf

thanks for that. I'll need to take a trip to the local electronics shop to pick up the resistor/diode as don't have spares so will try this on the weekend. I have an IR thermometer that can do continuous readings so might try that instead of my finger to check it out.

I've reseated the atmega chip and since doing that it appears to have stopped reconnecting over and over again. While initially looking good I'm not confident that's it so for now I'll monitoring and will try your suggestion this weekend.

cheers!

[BobW]

PAnother thought. Slit open an Ethernet cable and put the resistor in series with pin 6 - green/white I think and connect the amp to the controller. If you are powering with something with an amp or so of capacity, even a hard short in the amp shouldn't drag the +5 down on the controller. If the amp is running, you might be able to figure out what channel is causing the issue even if the device isn't going hard short continuously. Feed some noise into each channel and see if one is different / dead on the signal monitor.

[ShimmerNZ]

Thanks for your help Bob!

Managed to trace it to faulty component on Channel A. After replacing that she's been running good for 4+ days now.

[Dale.Reid]

What component did you find bad?

I worry about my Red and Blue, since there is no official service, certainly not one in this country, and these are very specialized boards.  I hope that a controller or complex chip doesn't go south, yet these are electronic systems and anything can happen.

So maybe it is whistling in the graveyard, but I was hoping you found something like a capacitor or some such rather than a more serious problem.
Thanks for sharing if you wish to do so. Dale

[BobW]

Thanks for your help Bob!

Managed to trace it to faulty component on Channel A. After replacing that she's been running good for 4+ days now.

Did you find it by limiting the current and looking for the "hot" culprit?
What actually was wrong?
Also, your station appears to only be sending data for one channel. 

Glad you made progress!

[ShimmerNZ]

So yeah, it seems I didn't fix it. Problem continued and finally today it seems to have died.

Discovery board lights up, LCD lights up, but eth has stopped working... no lights...nothing and lcd shows as eth not connected.

had to replace the 10uh inductor near the mini usb plug on the main board previously (it was getting hot from whatever was shorting).

Station 1243 is now down/dead.

Signed up for system blue. Too many weekends down trying to fix the current one (over estimated my ability to fix it).

[ShimmerNZ]

So yeah, it seems I didn't fix it. Problem continued and finally today it seems to have died.

Discovery board lights up, LCD lights up, but eth has stopped working... no lights...nothing and lcd shows as eth not connected.

had to replace the 10uh inductor near the mini usb plug on the main board previously (it was getting hot from whatever was shorting).

Station 1243 is now down/dead.

Signed up for system blue. Too many weekends down trying to fix the current one (over estimated my ability to fix it).

It's alive! Had to replace a second 10uh inductor to bring the Ethernet back to life. Root cause of where the short was occurring I finally managed to trace. It was the soldering to the Ethernet outlet on the amp. It looks like when I originally soldered it some of the solder bleed through to the other side of the board (but couldn't see it as the plug covers it). Not enough to short, but when any sideways twisting motion on the plug occurred it was enough to create the short. Despite the lan cable going through a strain connector there was still enough movement in the wind to cause the short on an irregular basis. Guess I have plenty of spare parts now =)

Hopefully no further issues and haven't broken anything in the process of trying to 'fix stuff' previously.