Announcement

Collapse
No announcement yet.

Update on USB Disconnect Issues (unable to detect USB modules during scan)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by Philbot View Post
    Our team does have a beak battery tester.

    Has MR issued a guideline for what constitutes an "acceptable" battery?

    In other words, if I have a battery that test "GOOD" on a beak, is that sufficient to expect reliable performance from the MR hardware? Or are there further requirements?

    Phil.
    And what about the teams that don't have one? Stretch our budgets thinner, be offered discounted battery replacements, or fix the bugs that did NOT plague the old system?
    FTC6460 mentor (software+computer vision+electronics), FPGA enthusiast. In favor of allowing custom electronics on FTC bots.
    Co-founder of ##ftc live chat for FTC programming--currently you may need to join and wait some time for help--volunteer basis only.

    Comment


    • I had commented earlier about how I got a new PDM that did not seem to be having problems. I would like to do an update on the condition of our PDM currently. Previously, almost any problem could be solved by restarting the robot on the robot controller or in rare cases power cycling for 5 seconds. Now, for most problems, I have to power cycle the robot for at least 20 to 30 seconds, and problems occur more often. Additionally, we have found that the electronics are not static-resistant, as in a recent practice we had a lot of lockups due to static from the metal robot touching the metal churro, and the robot would lock up and keep driving forward and then tip and roll of the ramp upside down.
      Programmer for Team 4997 Masquerade -- 2012 World Champions, 2014 - 2016 Division Finalists
      Founding Member of Team 6433 Neutrinos -- 2015 World Champions

      Check out my intro video to the new tech platform
      Check out my team's Robot Reveal for Res-Q

      Comment


      • Originally posted by Varun Singh View Post
        I had commented earlier about how I got a new PDM that did not seem to be having problems. I would like to do an update on the condition of our PDM currently. Previously, almost any problem could be solved by restarting the robot on the robot controller or in rare cases power cycling for 5 seconds. Now, for most problems, I have to power cycle the robot for at least 20 to 30 seconds, and problems occur more often. Additionally, we have found that the electronics are not static-resistant, as in a recent practice we had a lot of lockups due to static from the metal robot touching the metal churro, and the robot would lock up and keep driving forward and then tip and roll of the ramp upside down.
        Hi Varun,

        Thanks for the feedback. I'm sorry you are experiencing more issues.

        Would it be possible for you to retrieve the log files from your Robot Controller and Driver station and note the approximate times when certain key events (like the apparent static discharge event or other loss of control issues) occurred?

        Also, can you confirm which version of the FTC Driver Station and Robot Controller apps you are using (see Settings->About in both apps)?

        Tom

        Comment


        • You are right. By the time I got back to the lab to grab the blown PDM, I couldn't find it. Someone had already discovered the fuse and replaced it, I think. It's unfortunate that this is considered a solution to something like this, but at least it is a cheap fix. I think we are going to pre-emptively solve it by replacing all the PowerPole's on our motors withGrey and Orange housings so someone else on the team doesn't make the mistake at a critical time.

          Comment


          • Originally posted by FTC4106 View Post
            You are right. By the time I got back to the lab to grab the blown PDM, I couldn't find it. Someone had already discovered the fuse and replaced it, I think. It's unfortunate that this is considered a solution to something like this, but at least it is a cheap fix. I think we are going to pre-emptively solve it by replacing all the PowerPole's on our motors withGrey and Orange housings so someone else on the team doesn't make the mistake at a critical time.
            We color coded our motors by replacing the power pole plastic shells to different colors and use the same color electric tape to label the motor ports on the motor controllers. So hopefully, students will pay attention to match the color when plugging things in.

            Comment


            • In the old RC Car racing days, we would use things like car headlights to put a load on a battery. Check the voltage with a volt meter while the headlight is attached to the battery. It won't tell you the actual internal resistance of the battery, but... well, you should be able to weed out truly bad batteries. In my humble opinion, only a really pitiful 10 cell ( 12 volt ) battery , would drop bellow the 9 Volts needed under a normal robot load.


              Originally posted by hexafraction View Post
              And what about the teams that don't have one? Stretch our budgets thinner, be offered discounted battery replacements, or fix the bugs that did NOT plague the old system?

              Comment


              • Another option, would be to go find a local RC racing facility, or local hobby shop, and see if you can find someone that's been in the RC car/truck hobby for a while, ask if they know anyone with a Competition Electronics Turbo 30 charger/discharger . I just got mine out and while they were originally only setup to work with 7 cell or less battery packs, I just verified they will discharge 12 volt packs at anywhere from 1 to 25 amps ( could potentially go 30 amps, but it might damage the discharge circuits ) . These chargers were the top of the line for racers back in the day, but now are mostly obsolete because everyone runs Lithium Ion ( although you can still discharge the LithIon packs with the Turbo 30 as well. ) I bet if you could find one, someone would sell one for $25 or so... maybe even give it to you , if you told them what you were doing.

                Originally posted by GLDad View Post
                In the old RC Car racing days, we would use things like car headlights to put a load on a battery. Check the voltage with a volt meter while the headlight is attached to the battery. It won't tell you the actual internal resistance of the battery, but... well, you should be able to weed out truly bad batteries. In my humble opinion, only a really pitiful 10 cell ( 12 volt ) battery , would drop bellow the 9 Volts needed under a normal robot load.

                Comment


                • GLLad, I will look into trying to locate a Turbo 30 as you mentioned, or failing that, a large ohmic load.
                  FTC6460 mentor (software+computer vision+electronics), FPGA enthusiast. In favor of allowing custom electronics on FTC bots.
                  Co-founder of ##ftc live chat for FTC programming--currently you may need to join and wait some time for help--volunteer basis only.

                  Comment


                  • One of our coaches built one using some car headlights. The description and circuit are here:

                    http://www.tuxedopandas.org/ftc-battery-load-tester

                    Comment


                    • Before we owned a Battery Beak, we would go down to our local Batteries+ store and ask them to test our batteries for us. Tell them you are a local robotics team and they will normally do it for free.

                      - The Lazybotts

                      Comment


                      • We preach to our kids to operate with gracious professionalism. I hope FIRST and Modern Robotics also strives to operate in this manner. We are trying start a new platform for children to learn electronics that will last many years. There is potential with the electronics suite Modern Robotics has developed, but unfortunately we are running into beta test bugs. Some of these bugs are hardware related and in the short term would cost money to fix. In the long term this is going to create a much better program and platform for FIRST robotics. In gracious professionalism I would expect FIRST and Modern Robotics to admit there are hardware problems and fix them either free of charge or at a very reduced rate. Instead what I am seeing is defense of design flaws and blaming teams that they are using the product wrong. This is not gracious professionalism, this is profiteering. Please think about the long term goals of FIRST, and not get blinded by short term costs.

                        I would like to add that reverse voltage protection is a must in an electronic device for children learning electronics and wiring for the first time. The Tamaya connectors have been replaced by teams for years due to their short life span and ability to cause catastrophic failure at the worst times. I am not sure if FIRST and Modern Robotics ignored this and went with the Tamaya connector anyway, or Modern Robotics just didn't know their customer and didn't perform due diligence to see what teams wanted. Considering we paid $90 for a USB hub and power bus, I am disappointed reverse voltage isn't built in. The kids that have fried a PDM just feel horrible, and like they have let their team down. Just think when this happens at a competition, and it will happen a lot.

                        Comment


                        • The solution to the problems associated with modules not showing up during the scan even though everything is plugged in and set up correctly is actually because of the wires from modern robotics that connect from the core power distribution module to other modules. Our team encountered this problem, where out of 3 modules, only two were shown in the scan. We switched the wires around the modules and found out that a wire was defective. We got a spare wire (x5 in package), switched it with the defective wire, and then the problem disappeared. Do not bother checking battery voltage or anything else, because we changed batteries, checked all connections, restarted, all to no avail. I recommend buying all new wires at some nearby tech store, but only the ones that connect the core power distribution to other modules because only there are faulty, even though some may be working, they might fail on you later, even at crucial times. The wire quality is pretty bad, how could we get such cheap wires for a motor controller that's $80? I expect to get a high quality product for such a high price! Try getting all new wires, set them up, then scan it, I guarantee you it will scan all modules.

                          Comment


                          • Hello all,

                            I thought I'd post a result from continuing to look into the scan issue and batteries etc... I got a drone current sensor and hooked it up between a battery and a motor controller. I just got it wired tonight and it's pretty late, so just posting a bit. Will need to do a more thorough set of experiments later this week.

                            I measured delta I and delta V on a few different runs. Here are a few pieces of data from one example run:

                            Battery state according to Beak: 13.46 @ 0A, 13.363 @ 1A, 12.709 @ 5A, Rint = .163 and 86% charged. This is a weak / "bad" battery according to the Beak. However, we used this battery in scrimmages and competed with it last year. Robots ran ok, never noticed anything. Will try running it with the current monitor and an Arduino to log current/voltage while robot is do stuff.

                            So with one motor controller (4x470uF caps inside) I measured peak current draw of 2.3A and .427mS to get to 12v from power on. With no load it took .197mS, no surprise, you can see the effect from charging the caps, but it seems like less than a mS should be ok. Will check with a full board bot of controllers and other stuff, just testing the gear mostly tonight, seems to be working well.

                            Here is a screen grab if my Google drive share works...

                            https://drive.google.com/file/d/0ByL...ew?usp=sharing

                            Too early to draw any conclusions, but an interesting start to really understanding what power draw looks like on an FTC bot with the new electronics. It will also be interesting to see how the results match up with the MR results from their study!

                            More later,

                            Regards,
                            Bob
                            Bob France
                            --
                            Technical Mentor, Saratoga High School Engineering Club - GO FISH!
                            FRC 649
                            FTC 6165, 7390, 7641, 9965
                            TARC

                            Comment


                            • Well, I have to eat a bit of Humble Pie here in regard to the Competition Electronics Turbo 30 being useful for testing these 10 Cell NiMH batteries. It turns out, that it won't work, due to a voltage cutoff feature in the Turbo30. I had previously tested using a 3 cell LiPo , I had assumed because the Turbo 30 allowed me to discharge the LiPo ( which has a voltage of around 12 volts depending on state of charge ), I thought that the Turbo 30 didn't actually have any high voltage cutoff built in. I tested last night with my club's 10 Cell NiMH packs, and it won't allow you to start the discharge, due to exceeding some voltage threshold.

                              I am in the same boat really... the club really would rather not spend $100 on a device that only tests batteries, just doesn't seem like a valuable use of funds.

                              So, I'm now exploring another option... I found a 12 volt car heater, which consumes about 12 amps, powering it with a Robot Battery pack, measuring voltage over time, doing a basic plot of the discharge curve. I'll then evaluate the batteries condition based on this data. ( The little heater I bought for $16 at a local store department , in the automotive section )

                              Originally posted by hexafraction View Post
                              GLLad, I will look into trying to locate a Turbo 30 as you mentioned, or failing that, a large ohmic load.

                              Comment


                              • Can you clarify , exactly what your saying? Your saying that the main DC power wires ( the ones that connect with Anderson PowerPole connectors ) are somehow not reliably transferring the 12 volts ( and up to 10 amps or so ) to the various modules?

                                Originally posted by clock View Post
                                The solution to the problems associated with modules not showing up during the scan even though everything is plugged in and set up correctly is actually because of the wires from modern robotics that connect from the core power distribution module to other modules. Our team encountered this problem, where out of 3 modules, only two were shown in the scan. We switched the wires around the modules and found out that a wire was defective. We got a spare wire (x5 in package), switched it with the defective wire, and then the problem disappeared. Do not bother checking battery voltage or anything else, because we changed batteries, checked all connections, restarted, all to no avail. I recommend buying all new wires at some nearby tech store, but only the ones that connect the core power distribution to other modules because only there are faulty, even though some may be working, they might fail on you later, even at crucial times. The wire quality is pretty bad, how could we get such cheap wires for a motor controller that's $80? I expect to get a high quality product for such a high price! Try getting all new wires, set them up, then scan it, I guarantee you it will scan all modules.

                                Comment

                                Working...
                                X