by Mark1 » Tue Aug 02, 2011 8:42 pm
by Mark1
Tue Aug 02, 2011 8:42 pm
After about a week of a few hours of easy work per day, i finally managed to get the robot built to the point where i can program it to do stuff and it obeys.
Sadly i had to do it all alone, as the two friends that were supposed to help, turned out not to be interested enough, but the assembly was actually quite easy. Didn't even have to bother translating too much of the Jap. manual, except when it came to trim related sections (zeroing the servos, angle gauges, home position...).
I did have some problems tho - Sometimes, the tiny screws simply wouldn't go into the plastic part. The screwdriver would jump, and "eat" away at the head. I got worried i would damage a screw in a place where i couldn't get it out anymore, so i proceeded very carefully. I actually had to shorten many of the screws in places where i knew they wouldn't go in easily.
One plastic part was extremely nasty tho - the PCB base. At first a screw wouldn't go deep enough, got damaged, and I had to remove it with pliers, but that only broke off the head. I thought it was my fault, until later, when I was even more careful, a screw head broke off when I was tightening it with a screwdriver alone! Luckily that part isn't necessary for structural integrity, as it became impossible to put in another screw. That particular piece would have required some additional drilling first, to avoid this.
I managed to avoid any serious accidents, that would make it impossible to disassemble the robot and replace parts, but only cos I've been extremely careful.
I haven't actually started to cut the body covers yet, so it's naked for now, but programming it turned out to be fun enough already.
Unfortunately i am having some problems in this area as well:
My first problem is, that the
English version of H2H3 I found
isn't working properly. I managed to use it to zero the servos, and set the home position trim, and I could even write programs into RCB-3, but i simply can not make them run!
I was at the point where i thought the RCB-3 is malfunctioning, but then I tried the Japanese H2H3, and suddenly everything worked fine. Now I could write the programs to RCB-3 and execute them. But strangely, while it recognizes the programs I wrote in the English H2H3, they can not be executed for some reason.
Does anyone have a fully functional English H2H3, or knows what the problem could be?
For now i am forced to use the Japanese H2H3 for playing around. Luckily I remembered where the most important commands are, but it's annoying, besides I might miss something useful I could otherwise learn.
My second question relates to the ICS capabilities of the servos and the board. The servos and the board are supposed to provide a "learning" (or teaching) function, where you move the robot by hand, and take a snapshot of the positions, to save it as part of the program.
I tested it, and it actually works, except that
the ICS servo position read-out seems very inaccurate. When i pose the robot, and press the snapshot button, it jumps from the position it's in, to a slightly different one, as the servos power up and move to the values reported.
Is this normal, or should I be worried? Are these feedback inaccuracies typical for ICS servos, or is it just that these aren't accurate enough?
The error is sometimes a bit too big - the servo jumps by 5 or almost 10°.
For now I simply adjusted the values after taking the snapshot, but sometimes it's hard to get exactly to the position the robot was in, before the feedback error made it "jump".
Is there a way to correct this, or minimize the error, or should i just get used to it?
My last question is about KRS-4024 servo wear... As i was posing the robot, and adjusting the trim to make it balanced on one foot, i noticed that the knee servo was very hot afterwards.
Now i don't plan to have the robot standing on one foot all the time, but i am guessing that the leg and especially knee servos suffer a lot. This could become dangerous with time, if I spent too much time playing with trimming the robot just right, while it was standing on one leg...
What are the experiences of other MANOI owners here? How many servos did you destroy so far, and which ones? Do you think it is a good idea to upgrade to stronger ones?
Back to playing with the robot!
Thanks!
PS: One last question. Programming the robot while running on the battery is clearly not a good idea, if for no other reason, then simply because it doesn't last long enough to actually write any kind of decent program.
Does anyone know how much current this thing needs max? I want to give it an external power supply, before the battery drives me crazy. I'm hoping 5A at 12V will be enough?
After about a week of a few hours of easy work per day, i finally managed to get the robot built to the point where i can program it to do stuff and it obeys.
Sadly i had to do it all alone, as the two friends that were supposed to help, turned out not to be interested enough, but the assembly was actually quite easy. Didn't even have to bother translating too much of the Jap. manual, except when it came to trim related sections (zeroing the servos, angle gauges, home position...).
I did have some problems tho - Sometimes, the tiny screws simply wouldn't go into the plastic part. The screwdriver would jump, and "eat" away at the head. I got worried i would damage a screw in a place where i couldn't get it out anymore, so i proceeded very carefully. I actually had to shorten many of the screws in places where i knew they wouldn't go in easily.
One plastic part was extremely nasty tho - the PCB base. At first a screw wouldn't go deep enough, got damaged, and I had to remove it with pliers, but that only broke off the head. I thought it was my fault, until later, when I was even more careful, a screw head broke off when I was tightening it with a screwdriver alone! Luckily that part isn't necessary for structural integrity, as it became impossible to put in another screw. That particular piece would have required some additional drilling first, to avoid this.
I managed to avoid any serious accidents, that would make it impossible to disassemble the robot and replace parts, but only cos I've been extremely careful.
I haven't actually started to cut the body covers yet, so it's naked for now, but programming it turned out to be fun enough already.
Unfortunately i am having some problems in this area as well:
My first problem is, that the
English version of H2H3 I found
isn't working properly. I managed to use it to zero the servos, and set the home position trim, and I could even write programs into RCB-3, but i simply can not make them run!
I was at the point where i thought the RCB-3 is malfunctioning, but then I tried the Japanese H2H3, and suddenly everything worked fine. Now I could write the programs to RCB-3 and execute them. But strangely, while it recognizes the programs I wrote in the English H2H3, they can not be executed for some reason.
Does anyone have a fully functional English H2H3, or knows what the problem could be?
For now i am forced to use the Japanese H2H3 for playing around. Luckily I remembered where the most important commands are, but it's annoying, besides I might miss something useful I could otherwise learn.
My second question relates to the ICS capabilities of the servos and the board. The servos and the board are supposed to provide a "learning" (or teaching) function, where you move the robot by hand, and take a snapshot of the positions, to save it as part of the program.
I tested it, and it actually works, except that
the ICS servo position read-out seems very inaccurate. When i pose the robot, and press the snapshot button, it jumps from the position it's in, to a slightly different one, as the servos power up and move to the values reported.
Is this normal, or should I be worried? Are these feedback inaccuracies typical for ICS servos, or is it just that these aren't accurate enough?
The error is sometimes a bit too big - the servo jumps by 5 or almost 10°.
For now I simply adjusted the values after taking the snapshot, but sometimes it's hard to get exactly to the position the robot was in, before the feedback error made it "jump".
Is there a way to correct this, or minimize the error, or should i just get used to it?
My last question is about KRS-4024 servo wear... As i was posing the robot, and adjusting the trim to make it balanced on one foot, i noticed that the knee servo was very hot afterwards.
Now i don't plan to have the robot standing on one foot all the time, but i am guessing that the leg and especially knee servos suffer a lot. This could become dangerous with time, if I spent too much time playing with trimming the robot just right, while it was standing on one leg...
What are the experiences of other MANOI owners here? How many servos did you destroy so far, and which ones? Do you think it is a good idea to upgrade to stronger ones?
Back to playing with the robot!
Thanks!
PS: One last question. Programming the robot while running on the battery is clearly not a good idea, if for no other reason, then simply because it doesn't last long enough to actually write any kind of decent program.
Does anyone know how much current this thing needs max? I want to give it an external power supply, before the battery drives me crazy. I'm hoping 5A at 12V will be enough?