Hello everyone!
The power switch of our magni robot is not working - the raspberry is always on.
The motor switch works as usual and both Blue and Red S/W too.
I was hopping you could help us do the troubleshooting or advise us on what should we do.
Thank you in advance!
Please identify the version of the main board per this page:
Also on same page identify the switch board you have (at end of the page)
When you say Blue and Red switch do you mean that when the left switch (black top) is out the Blue led is on on the switch board and when you push in the black switch the blue led goes off?
What happens to the row of 5 blue leds on main board when the blue switch is âoutâ and the blue led is off?
Thank you,
Mark
Hello,
I have the version 5.1 of the main board and the version 2.0 of the switch board.
When I press down the black switch the blue led lights up and when is out the blue led goes off (I think is the normal behaviour).
But the 5 blue leds never go off, regardless the black switch position. And the raspberry does not turn off either. All leds only turn off when I unplug the batteries.
Thank you so much for the availability!
Good info. Next few questions are these:
Is it true that the single led off on the lower left of the MCB board always on as well as the 5 leds and Pi? I think your answer will be âyesâ but it is important to be sure. This led is one led to the left of where the switch board 14 pin connector is. You cannot miss it if it is on, it is bright. It has a label of âLED1_ECB_POWâ as well.
Do you use any of the 5V or 12V auxillary power or main power the board offers? The main connectors for this power is the two white 4-pin big connectors at the top of the board labeled with âMAINâ or âAUXâ (both of these offer user power where we prefer use of âAUXâ first by customers.
Since your board is rev 5.1 can you tell me if the large plug in fuses are green or purple which helps me know what order of boards this was from. Better yet if you have a white bar code tag on the board can you tell me the full number because that identifies the date the board was made exactly. I have seen many of the boards with purple fuse donât have the tag which was a production oversight.
Thanks,
Mark
I forgot to ask if the switch ever worked for a while THEN stopped working or has always been bad?
Also please try to take a picture with all power OFF the best you can do of all the area in the lower left of the board. I want to check and look for issues with parts or contamination on the board.
If you can remove the switch board for this picture it will allow seeing the area much better.
You can turn off power and REMOVE RED BATTERY CABLE. Then unscrew 2 screws for the switch board and the switch board can then pull out of the connector on the MCB for better picture. Take a couple pictures and one picture showing if you have any custom cables hooked to the MCB.
Thanks,
Mark
Hello Mark,
So according to your questions:
These pictures were taken following your instructions: unplugging the batteries and with the switch board pulled out. If you need more info let me know.
Once again thank you very much!
I had tried to send you a message from this forum but got a message that that feature is disabled. I will look into that. In the mean time we are going to add an email for our support issues so details for getting a board fixed can be discussed offline. The email would go to people doing repairs so it would be an easier process. It is a holiday but we hope to get that setup soon, perhaps this week.
If this is a very urgent issue please say so and I will get you other information for contact information on one of our direct emails.
Sorry you are seeing this problem and please stand by
Mark
Hello
I had the same issue. The power swich stop working. The robot is on when batteries are connected. But in my case it stoped working after a short circuit between pins 49 and 50 on the sonar board. There aren`t any visible damage on the main board. The version of the main board is 5.2 and the swich board is 2.2. Can you help me?
Thanks!
Are you saying your board works now but is stuck always on ever since an accidental short of pin 49 and 50? I am very interested in your situation because it may add insight to a fairly new issue with main power on rev 5.2 boards. Thank you for your post I think it will be very helpful and we will get your issue sorted out too.
Inspired by your comment on the shorting of pin 50 to pin 49 I re-visited some tests I had done on rev 5.2 and setup a âtorture testâ to see if I could re-create a way to make the rev 5.2 main power ECB fail (that is what you have by the way). We made a change for safety reasons to put the battery voltage that shows up on the 50 pin connector behind the main ECB. It seems now that this leads to this failure situation.
This applies ONLY to rev 5.2 and there is no need on rev 5.1 or 5.3 coming soon.
So I repeated some âtorture testsâ and this time tried many times to do very fast shorts and long shorts and sure enough I was able to make the ECB fail after 15 or so tries.
We had changed this already for our rev 5.3 board but rev 5.2 has the very large R706 soldered down to the board. So for those people with rev 5.2 boards AND the ability to remove a large SMT resistor with a SMT rework heat gun here is my suggestion to avoid possible failures later due to unexpected shorting of the power on the 50 pin connector.
There are other ways to do the effective removal of the effect of R706 which I only suggest if you are SAFE and good with an X-Acto knive or dremel tool diamond wheel cutter (one of my all time favorite PC mod tools FYI). If you have little to no experience in trace cutting, DO NOT TRY THIS FOR YOUR OWN SAFETY! An Exacto knive is VERY dangerous and use âboy scoutâ cut away from your fingers if you go this route!
To Cut Trace: You will see a very fat topside trace that goes from R706 up the left edge of the board. If you cut this fat trace on the top of the PCB right above the âRâ in the white silkscreen of âRev 5.2â and be sure what you have done does NOT short this trace to the very nearby ground plane or cut much past the trace itself into the board that will do it.
A mod like this will possibly save you from a later failure if you accidentally short by dropping screw or something that shorts this battery power.
We are going to have this addressed on rev 5.3 boards hitting production very soon.
Hello
Thank you for replying so quickly!
there is another thing that is worrying me, could the charging circuit of the robot be damaged? Since I had the short circuit I haven`t tried to charge it.
In a near future I would like to charge the robot throught some contact pads like this : https://www.roboteq.com/all-products/robopads-charge-system
Do the robot have some dedicated pins where it could be attached?
There is a docking bay in development but I donât know itâs release date.
The failure impacted the power MOSFET and my thought would be that would not break the charger.
Connect it up with MCB being off and then power up your charger while noting with a current probe if there is excessive current from the charger (over 5 amps for our chargers) or if the charger voltage sags in a huge way if you donât have current measurement ability.
The separate contacts is not a bad idea at all. We do not bring out the same points on a easily accessible connector yet and I will think on having one next time around.
What you could do is use a modified switchboard where the large XLR jack is taken off and you solder to the two connections the charger goes into now. OR easiest is have your own XLR plugged into our jack and that then goes down to your robopad connection. This of course has itâs downfalls but would be âcleanâ.
Iâm experiencing an issue similar to what was described here. In my case I was inserting a USB mouse into the RaspberryPi USB port. The narrow space under/behind the front âbeamâ of the Magni chassis makes this a little difficult and the cord slipped out of my grasp and a metal part of the plug touched something either on the MCB or the switch board â got a spark! Everything with respect to the RaspberryPi has continued to work as normal and the motor control seems to be working fine (it behaves as expected with teleop_twist_keyboard but Iâve discovered that now I canât fully âturn offâ the MCB with use of the black main power button on the switch board. When the button is depressed the blue main power LED on the switch board turns off, but there are lights that remain illuminated on the MCB (LED 11 indicating 24VM and two other LEDs in the strip of four just above where the switch board connects to the main board â 12VM and 5 VA). The only way to fully cut power seems to be detaching a cord at the battery spade connector. Now it seems that the batteries wonât charge either. Itâs a version 5.3 MCB. Any tips for troubleshooting or repair?
The short reply is: Please send an email to support@ubiquityrobotics.com and in the subject please state something like âFailed power on MCB rev 5.3â then we can take this up from here on using email.
The more detailed reply is: We control the main power with something we call an Electronic Circuit Breaker or âECBâ. What you are describing, murpht2, is that the main power ECB has a failure that keeps main power active no matter the state of the large Blue led on the switch board. The other thing you are describing is that 2 out of 4 of the power supplies that âfollowâ the main power are non-functional. Our version 5.3 MCB uses a light conformal coating to help prevent shorts from happening but they still can happen if metal parts short the electronics on the MCB. We recommend fully disconnecting power for cable changes but we do understand how it is very normal for us and customers to give it their best shot at plugging in USB or the ethernet. Earlier boards were even worse as there were jacks and things sticking up in that area so I designed them out but dropping the connector on the board can short out and cause electrical failure.
Stay Healthy, Mark
Thanks, Mark, following up via email now.
