No continuous charging anymore

[cdondrup]

I just observed a weird behaviour of the charging station:

Today I found our robot sitting on the station with a dead battery. The key was set to "on" so it must have been charging while on the station. I switched the robot on and the charging immediately began. The only problem was, that the charging was not continuous but the LED was blinking in 1 second intervals like when the battery is at 100% (see issue #26 and this link for a video: http://youtu.be/IW-odD6zXOk). The answer to our question if this behaviour is intended was: @ml-stu "Yes, it is right when the embedded pc is shut down ant the batteries are at 100% charge." But we still observed this behaviour even when the embedded PC and our version of the mira bridge was running.

Now the station seems to do this all the time even if the battery is at 0%. The voltage alternates between 3 and -1 so it does not really charge even if the LED is on.

@ml-stu @chmartin21 any ideas what could cause this and how to fix it?

[cdondrup]

Charging using a cable and the rear power socket works fine, though.

[ml-stu]

after you changed all the electronics in February (and the electronics, you sent back do their job here without problems) it seems there is a problem with the cabling (contact resistance too high). To verify this could you please record all can messages for about a minute while the robot is at the docking station and shows the blinking charging LED behaviour? You will find a "Recorder" button at the bottom of the MIRA center. Drag the "incomingCAN" channel to the recorder view and start recording. Please send us the file per email or upload it here.

[cdondrup]

@ml-stu we just recorded the incomingCAN and also everything that was in the charging category. You can find the files here: http://lcas.lincoln.ac.uk/owncloud/public.php?service=files&t=42267cb461b949e9341601cd594ba129 There are two files:

• mira_can_recording.tape contains the incomingCAN
• full_can_mira_log.tape contains the incoming and outgoing CAN and the three topics in the charging category

Since we have an additional robot here in Lincoln at the moment we tried to put it on the charging station and it showed the same behaviour.

[Jailander]

@ml-stu @chmartin21 Hi do you have any news on this? we really need the charging station functional soon.

[cdondrup]

Any news @ml-stu @chmartin21 ?

[chmartin21]

Sorry, we're still investigating this issue. Probably, you will get an answer beginning of next week.

[ml-stu]

We assume a defect at the line filter or the primary side cabling. Please send us the charging station for testing and repairing to Ilmenau.

[cdondrup]

The whole station or just the electronics?

[ml-stu]

The whole station. We need the complete electrical path for testing.

[cdondrup]

Ok, we will pack it up and send it to you asap. Thank you very much for your help so far.

[ml-stu]

The charging station arrived at Ilmenau, we checked it up and it shows the same mysterious behaviour. There was a defect IC at the small PCB so we installed a new PCB, tested the charging station (works fine) and now it is on its way back to you.

[cdondrup]

That's great news. Thank you very much.

[ml-stu]

ok, it's back at Lincoln. Maybe you have to ask at your post room (Andy Barton) http://nolp.dhl.de/nextt-online-public/set_identcodes.do?lang=en&idc=968536250278

[cdondrup]

We picked it up after lunch. Works like a charm. Thank you very much for the quick repair!

[cdondrup]

Sadly, we seem to have the same error again. I'll post the email conversations for documentation.

Hi Christians,

since wednesday night, Linda's charging station seem to be malfunctioning again. The symptoms are the same as described here: https://github.com/strands-project/scitos_robot/issues/43.

@CM: As far as I remember, the problem was caused by a malfunctioning IC on the charging station PCB. Last time, we have reported the error on April and the replacement arrived on June.

Therefore:

1) How likely is the error caused by the same reason ?
2) Can we try to test the electrical paths and the station's PCB ourselves ?
3) If the reason is the same as before, could you send us two replacement PCBs so our data collection is not interrupted ?

@CD: Could you reopen the issue on github ?

cheers,

Tom 

Hello,

I heard the charging station is malfunctioning again. Could you please confirm these points:

* Linda is not charging at the charging station but charging with the AC power cord works fine

* powering off the charging station (mains cable) and on again does not solve the problem

* when plugging mains at the charging station all LEDs (green, red, yellow) turn on for about a second and only green remains on

Kind regards

Stefan Ulbrich 

Hi Stefan,

To answer your questions:

>
> I heard the charging station is malfunctioning again. Could you please confirm these points:
>
>
> * Linda is not charging at the charging station but charging with the AC power cord works fine
>
Charging with the power cord works fine and shows these stats:

voltage: 26.9700012207
current: -8.92200088501
lifePercent: 97
lifeTime: -1
charging: True
powerSupplyPresent: True
cellVoltage: [3.3450002670288086, 3.372000217437744, 3.379000186920166, 3.373000144958496, 3.376000165939331, 3.38100004196167, 3.373000144958496, 3.373000144958496]

The first cell seems to be a bit lower (0.03) than the others. Is that a problem?

> * powering off the charging station (mains cable) and on again does not solve the problem
>
Sadly, it doesn't. Unplugged it until the green LED went off (+1 minute) and plugged it back in.

> * when plugging mains at the charging station all LEDs (green, red, yellow) turn on for about a second and only green remains on
>
Yes, it shows that behaviour. As soon as we put the robot on the station the charging LED starts blinking with ~0.5hz. When it's off it shows normal discharging:

voltage: 26.7820014954
current: 4.33500003815
lifePercent: 97
lifeTime: -1
charging: False
powerSupplyPresent: False
cellVoltage: [3.32200026512146, 3.3480000495910645, 3.3550002574920654, 3.3540000915527344, 3.3510000705718994, 3.3510000705718994, 3.3520002365112305, 3.3490002155303955]

But when it's on it only charges for a fraction of a second with a very low current:

voltage: 26.8680019379
current: -0.866000056267
lifePercent: 97
lifeTime: -1
charging: True
powerSupplyPresent: True
cellVoltage: [3.3310000896453857, 3.3580002784729004, 3.366000175476074, 3.3630001544952393, 3.3630001544952393, 3.3630001544952393, 3.3610000610351562, 3.3580002784729004]

Best wishes,
Christian 

The error is exactly the same as shown in above video.

[ml-stu]

We will send you two controller boards for the charging station.

[ml-stu]

The two boards will arrive today: http://nolp.dhl.de/nextt-online-public/set_identcodes.do?lang=en&idc=968530932448&rfn=&extendedSearch=true

[gestom]

I have replaced the controller board and let the robot to charge itself on the station. Everything seems to be fine. Thank you for the boards.

[cdondrup]

The new board broke after almost exactly one month. We are experiencing the issue again. We will try and replace the board with the second one you sent us but we suspect that there might be something else frying the board quite regularly. How can we investigate this?

[ml-stu]

Maybe it is an ESD problem. Do you drive on carpet? Did you experience ESDs when touching the robot?

[cdondrup]

There is a lot of ESD going around the office lately. The robot is driving on carpet and sometimes you do get a shock when touching it. Is there some way to prevent this from frying the board?

[ml-stu]

1) Please check that the left charging plate is connected to earth (charging stations metal case). We do not know if this is already done in your charging station. If not you could wire from the contact plates screw to the earth screw or stick a copper tape from the plates end to the line filter.

2) Isolate the first 10-20mm of the right charging plate so the docking robot will always contact the left plate at first.

[cdondrup]

We currently haven an experiment running for which we need the station desperately. We will start by using insulating tape on the right plate but we are unsure if the left plate is connected to the casing. We can have a look on Sunday and have to hope that the board survives. Thank you very much for the tip!

[gestom]

Which earth ? The protective earth (PE) or the neutral (N) ?

[ml-stu]

PE

[ml-stu]

Is the charging station working now with the modifications? Can you please send the defective boards back to MetraLabs for repair?

[ml-stu]

@gestom, @cdondrup: Is the charging station working now with the modifications? Can you please send the defective boards back to MetraLabs for repair?