Comments for Bot Bench http://botbench.com/blog I'd Rather Be Building Robots Fri, 18 May 2012 21:01:43 +0000 hourly 1 http://wordpress.org/?v=3.3.2 Comment on Connecting the NXT to an ArduinoMega by David Fort http://botbench.com/blog/2009/08/04/connecting-the-nxt-to-an-arduinomega/#comment-2076 David Fort Fri, 18 May 2012 21:01:43 +0000 http://mightor.wordpress.com/2009/08/04/connecting-the-nxt-to-an-arduinomega/#comment-2076 Just confirming: On the Rev 3 Arduino Mega 2560 R3 "A" and "D" pins (1,8,4,5) of the Resister Network RN1 are not connected to anything (according to the board and schematic files on the Hardware page of the Arduino web site (which I opened with Eagle Lite version 6)). Therefore, it is safe to completely remove RN1 rather than try to saw it apart or whatever. We tried several techniques, and the easiest for us was a pair of Weller tweezers, held upside down to heat all 8 pins at the same time. Just confirming:
On the Rev 3 Arduino Mega 2560 R3 “A” and “D” pins (1,8,4,5) of the Resister Network RN1 are not connected to anything (according to the board and schematic files on the Hardware page of the Arduino web site (which I opened with Eagle Lite version 6)).

Therefore, it is safe to completely remove RN1 rather than try to saw it apart or whatever. We tried several techniques, and the easiest for us was a pair of Weller tweezers, held upside down to heat all 8 pins at the same time.

]]> Comment on Connecting the NXT to an ArduinoMega by David Fort http://botbench.com/blog/2009/08/04/connecting-the-nxt-to-an-arduinomega/#comment-2075 David Fort Fri, 18 May 2012 17:53:19 +0000 http://mightor.wordpress.com/2009/08/04/connecting-the-nxt-to-an-arduinomega/#comment-2075 Thank you Xander. Once again you have saved us! Like you, we have been trying to get the NXT to talk I2C to the Arduino Mega 2560 (ours is R3) and it works with other devices, etc., but on the oscilloscope we saw that signal was not pulled down as far as we expected and from the current measurements, we concluded the pullup must only be about 8K, not the 20K to 50K as in the atmel datasheet, but looking at the schematic we did not notice the additional external pullups because they aren't connected via lines. On the R3 board, they are in RN1 (Resister Network 1) and they are on the middle 2 pins of that part (which has 4 10K resisters in one package), and as far as I can tell by looking at the pads and by searching the .pdf of the schematic for "RN1", the top and bottom resisters in the network are not used for anything. RN1 is in the same general area as the 2 discrete resisters you have circled in your picture (of the older revision of this board). We removed the whole thing and it seems to still be working, but we aren't using anything other than I2C (aka TWI) right now, so if those other resisters really are needed for something, it might bite us later. They don't seem to be. One really puzzling puzzle, when we just "read the pin" values, we were able to see the NXT signal in the Arduino firmware. That led us to think the signal was "low enough" even though not all the way to 0V, but after removing the resister network, TWI/I2C actually WORKS with the NXT! We think this is because the signal path to the "read pin" is different than the signal path to the "SCL INPUT" (the former goes through a schmitt trigger and a synchronizer), so maybe that is why you can get the right value from the port, but the TWI interface doesn't see it. (fwiw, the I2C spec says the "low" signal needs to be 0.3*V where V is the bus voltage, 4.3 or 5 in this case. Thus, it should be less than 1.5V to be "sure", and ours was a little more than 2V when "off", so clearly outside the acceptable spec, even though "almost working"). Again, Thank you for this post. Thank you Xander.
Once again you have saved us!

Like you, we have been trying to get the NXT to talk I2C to the Arduino Mega 2560 (ours is R3) and it works with other devices, etc., but on the oscilloscope we saw that signal was not pulled down as far as we expected and from the current measurements, we concluded the pullup must only be about 8K, not the 20K to 50K as in the atmel datasheet, but looking at the schematic we did not notice the additional external pullups because they aren’t connected via lines.

On the R3 board, they are in RN1 (Resister Network 1) and they are on the middle 2 pins of that part (which has 4 10K resisters in one package), and as far as I can tell by looking at the pads and by searching the .pdf of the schematic for “RN1″, the top and bottom resisters in the network are not used for anything.

RN1 is in the same general area as the 2 discrete resisters you have circled in your picture (of the older revision of this board).

We removed the whole thing and it seems to still be working, but we aren’t using anything other than I2C (aka TWI) right now, so if those other resisters really are needed for something, it might bite us later. They don’t seem to be.

One really puzzling puzzle, when we just “read the pin” values, we were able to see the NXT signal in the Arduino firmware. That led us to think the signal was “low enough” even though not all the way to 0V, but after removing the resister network, TWI/I2C actually WORKS with the NXT!

We think this is because the signal path to the “read pin” is different than the signal path to the “SCL INPUT” (the former goes through a schmitt trigger and a synchronizer), so maybe that is why you can get the right value from the port, but the TWI interface doesn’t see it.

(fwiw, the I2C spec says the “low” signal needs to be 0.3*V where V is the bus voltage, 4.3 or 5 in this case. Thus, it should be less than 1.5V to be “sure”, and ours was a little more than 2V when “off”, so clearly outside the acceptable spec, even though “almost working”).

Again, Thank you for this post.

]]> Comment on ROBOTC: Using the NXT 2.0 Colour Sensor by kelso http://botbench.com/blog/2011/02/16/robotc-using-the-nxt-2-0-colour-sensor/#comment-2074 kelso Thu, 17 May 2012 11:18:47 +0000 https://mightor.wordpress.com/?p=916#comment-2074 yes, i ruduced the amount of light allowed to get to the light. yes, i ruduced the amount of light allowed to get to the light.

]]> Comment on ROBOTC: Using the NXT 2.0 Colour Sensor by Xander http://botbench.com/blog/2011/02/16/robotc-using-the-nxt-2-0-colour-sensor/#comment-2072 Xander Wed, 16 May 2012 12:36:47 +0000 https://mightor.wordpress.com/?p=916#comment-2072 There is not an awful lot I can do for you, unfortunately. The sensor either works or it does not. I have mixed results with this sensor. It is most likely not the code you have, it's just a flaky sensor. Did you try it in a different room with different lights? There is not an awful lot I can do for you, unfortunately. The sensor either works or it does not. I have mixed results with this sensor. It is most likely not the code you have, it’s just a flaky sensor. Did you try it in a different room with different lights?

]]> Comment on ROBOTC: Using the NXT 2.0 Colour Sensor by Kelso Watsey http://botbench.com/blog/2011/02/16/robotc-using-the-nxt-2-0-colour-sensor/#comment-2069 Kelso Watsey Wed, 16 May 2012 11:29:14 +0000 https://mightor.wordpress.com/?p=916#comment-2069 Y you no help Y you no help

]]> Comment on ROBOTC For Arduino is Out Now! by Xander http://botbench.com/blog/2012/05/11/robotc-for-arduino-is-out-now/#comment-2068 Xander Wed, 16 May 2012 06:22:20 +0000 http://botbench.com/blog/?p=2249#comment-2068 John, Th reason the Uno does not support debugWriteStream is because it does not have string support due to memory limitations. The ROBOTC developers are looking into a way to support a simplified debugWriteStream without the strings for the smaller memory Arduino platforms. John,
Th reason the Uno does not support debugWriteStream is because it does not have string support due to memory limitations. The ROBOTC developers are looking into a way to support a simplified debugWriteStream without the strings for the smaller memory Arduino platforms.

]]> Comment on Exposed: HiTechnic Angle Sensor by Xander http://botbench.com/blog/2010/10/28/exposed-hitechnic-angle-sensor/#comment-2067 Xander Wed, 16 May 2012 05:29:26 +0000 https://mightor.wordpress.com/2010/10/28/exposed-hitechnic-angle-sensor/#comment-2067 Sure, you can measure RPM (revolutions per minute) with it. If you know how fast something spins, you could conceivably calculate speed. - Xander Sure, you can measure RPM (revolutions per minute) with it. If you know how fast something spins, you could conceivably calculate speed.

- Xander

]]> Comment on Exposed: HiTechnic Angle Sensor by Chris http://botbench.com/blog/2010/10/28/exposed-hitechnic-angle-sensor/#comment-2066 Chris Wed, 16 May 2012 01:48:33 +0000 https://mightor.wordpress.com/2010/10/28/exposed-hitechnic-angle-sensor/#comment-2066 Can this be used as a tachometer? For speed measurement? Can this be used as a tachometer? For speed measurement?

]]> Comment on Building Robots At Schuberg Philis by Xander http://botbench.com/blog/2012/05/13/building-robots-at-schuberg-philis/#comment-2064 Xander Mon, 14 May 2012 16:26:45 +0000 http://botbench.com/blog/?p=2267#comment-2064 The rocky start was a tongue-in-cheek remark where we were making fun of Nextstorm's robot building abilities, which are, in fact, pretty amazing :) The rocky start was a tongue-in-cheek remark where we were making fun of Nextstorm’s robot building abilities, which are, in fact, pretty amazing :)

]]> Comment on Building Robots At Schuberg Philis by GRK_Astronomer http://botbench.com/blog/2012/05/13/building-robots-at-schuberg-philis/#comment-2063 GRK_Astronomer Mon, 14 May 2012 15:43:57 +0000 http://botbench.com/blog/?p=2267#comment-2063 Rebounding from "All work and no play ...," Schuberg Philis breaks out of the box and enters into the world of "mission critical" LEGO Mindstorms Robotics. I believe that this was one of the first critical steps of moving LEGO Mindstorms from the toy stores and into the boardroom. I have long thought that Mindstorms could be a valuable tool in teaching corporate HR courses to management, like the "Building Bridges" courses of the 1980s. The sets are on the table, now all we need is a clever HR consultant to come in and build some management courses using Mindstorms to teach managers as well as letting them have some fun on occasions. Well done, even of the start was rocky. Rebounding from “All work and no play …,” Schuberg Philis breaks out of the box and enters into the world of “mission critical” LEGO Mindstorms Robotics. I believe that this was one of the first critical steps of moving LEGO Mindstorms from the toy stores and into the boardroom. I have long thought that Mindstorms could be a valuable tool in teaching corporate HR courses to management, like the “Building Bridges” courses of the 1980s. The sets are on the table, now all we need is a clever HR consultant to come in and build some management courses using Mindstorms to teach managers as well as letting them have some fun on occasions. Well done, even of the start was rocky.

]]>