Pololu Blog » User Profile: Paul »
Comments by Paul
Home | Forum | Blog | Support | Ordering Information | Wish Lists | Distributors | BIG Order Form | About | Contact
© 2001–2024 Pololu Corporation
Home | Forum | Blog | Support | Ordering Information | Wish Lists | Distributors | BIG Order Form | About | Contact
© 2001–2024 Pololu Corporation
|
Log In
|
My Account
|
Log Out
|
Shopping Cart
US toll free: 1-877-7-POLOLU ~
(702) 262-6648 |
|||||||||||||||
| Catalog | Forum | Blog | Support | Ordering | Distributors | About | Contact | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
How to make a Balboa robot balance, part 5: popping up and driving around
- 18 March 2024Hello,
Sounds great that your robot is almost balancing!
In Balance.h there is a description about each constant that says how to adjust it. I think it makes sense to go through these basically from the top to bottom and follow the instructions. Can you let me know if you are getting stuck somewhere in particular?
-Paul
How to make a Balboa robot balance, part 5: popping up and driving around
- 13 March 2024Hello Nir, I'm glad you enjoyed this series.
If you want to use our example code as a starting point, then yes, changing GEAR_RATIO is a good start. Of course other values might need to be calibrated too.
To flip the motor direction, you can use the flipLeftMotor() and flipRightMotor() functions in the Balboa library. You should also correspondingly flip the encoder readings, maybe right where getCountsLeft() and getCountsRight() are called.
It's probably a good idea to test these changes individually; make the robot drive forward and verify that it really goes the direction you expect and the encoder counts increment the way you expect. Then check the directions in the gyro and accelerometer too. Otherwise a single sign error might cause the behavior to seem totally wrong, and you'll have no idea where the error lies.
I would be interested to see pictures of how you are mounting the wheels directly to the motor shafts! Please note that the plastic gearbox is there partly to isolate the motor from physical shocks involved in balancing, especially when popping up or falling down; if you connect them directly you might risk damage to the metal gears.
-Paul
Taking control of a typosquatting domain with a UDRP case
- 11 April 2022Thanks for the feedback. The article seems to be talking about a way to efficiently identify and block phishing domains, for example within a few hours after they are registered. But we don't get to block domains for our customers, and I don't know what else we can easily do. For example, polulu.com was not (to our knowledge) engaged in phishing, so I don't think a takedown request to the registrar would have been effective. I'd be happy to hear about specific approaches to taking down a domain that are easier than UDRP!
How to make a Balboa robot balance, part 5: popping up and driving around
- 10 November 2020Hello, John.
You should be fine with the 50:1 motors and 80mm wheels that I used for this example, but there are many configurations that will work, especially considering that the bumper cage can help it start out with a boost toward vertical.
-Paul
How to make a Balboa robot balance, part 5: popping up and driving around
- 3 March 2020Hello, Ali.
I'm glad you are enjoying my posts! The reason to adjust
speedLeftandspeedRightis that the main balancing algorithm tries to get all of these variables to zero. If you don't adjust the speeds, according to the formula formotorSpeed, something else is going to have to change to compensate; I think it will be slower to start moving and never catch up to where it's supposed to be.-Paul
New product: VL6180X Time-of-Flight Distance Sensor Carrier
- 14 February 2018Hello,
I'm glad to hear that you like the parts and our prices!
It also sounds impossible to me, and we don't have detailed information about how the ST devices work, but they do specifically talk about measuring the time of flight, and datasheets for similar parts like the ISL29501 give more details about how you can use analog signal processing to determine an absolute flight time. Of course the electronics can't do much in a few picoseconds, but signal rise times of 10 ns/V and millivolt sensitivity are normal, which gets you to that timescale. Building an interferometer in a chip and sweeping the frequency of a laser diode sounds a lot harder.
By the way, the light has to go there and back, so you can double your times, which helps a bit.
-Paul
How to make a Balboa robot balance, part 4: a balancing algorithm
- 12 December 2017Hello,
Part 5 of this series discusses driving around. Please also see the comments for some more suggestions.
-Paul
How to make a Balboa robot balance, part 5: popping up and driving around
- 5 December 2017Hello,
I don't have any immediate plans to do more tutorials. But if you have made it this far with your Balboa, you should be ready to explore these ideas on your own! Any of these ideas would be taking it further than what I have done, so I would be interested to hear what you try and comment on it; maybe you could post your progress to our forum. For navigation, you might be interested in my dead reckoning robot from a few years ago; you could probably use the same basic approach to do more precise navigation with Balboa. There are a ton of tutorials available for IR remote control; since the Balboa doesn't have any built-in support for this, just look for any Arduino remote control project. Here's an example from our Zumo library using the built-in sensors on the Zumo, which are also available as discrete modules (though there are other sensors that are more appropriate for remote control applications).
-Paul
How to make a Balboa robot balance, part 2: inertial sensors
- 27 March 2017That's a reasonable idea. I have never actually seen it fail to initialize the sensor. If it did occasionally fail, I might change that line as you suggested, but only after looking into the source of the problem and determining that there is nothing I can do about it. Our examples are not intended to be robust against communication failure, so if you need something more robust, you might also want to look into re-initializing the sensor if it fails during loop() and consider enabling the AVR's watchdog timer.
Building a Raspberry Pi robot with the A-Star 32U4 Robot Controller
- 4 October 2016Thanks for the comment! I have updated the command as you suggested.
-Paul