by alejandrocerrano » Tue Nov 24, 2015 11:41 pm
by alejandrocerrano
Tue Nov 24, 2015 11:41 pm
Hi, VMatos.
Thanks for your email.
I am taking an Specialty on Embedded Systems, at University. My strongest skills are in C and C++, but at medium-high level programming.
Embedded Systems was always a pending subject for me. I am learning.
Well, my Thesis Director gave me a Robonova I, which is not working (seems some students damaged the motherboard some time ago).
So, I have 2 goals (as per my Thesis Director):
1. Having the Robonova walking again.
2. Having the Robonova processing images and "taking decisions" accordingly.
As I need to use University economic resources (this is a good practice to have students working in order way), I need to provide reasonable arguments before buying hardware, for instance.
1. So, for the first goal, the logical first step is to prove that the original motherboard is working / damaged.
Once proved, I have to suggest alternative motherboards (in case of damage) to replace it.
One alternative is buying the natural replacement for 3024 Mother, which is the 3024FX released by the same Korean manufacturer (HITEC provided the name of the real manufacturer to me; I contacted them, and they told me that 3024 mother and 3000 mother have been discontinued because they caused damages to the servos. Instead, they have release a FIX, the 3024FX, which is used by the Robonova II - Metal Fighter).
I appreciate your suggestion of Arduino-like for 24 servos and Dagu Spider also. Regarding Arduino, I suppose you are talking about some Arduino shield, which I did not know before your mention. I have also checked the Dagu Spider controller and it seems also a very good option.
BUT, look at this: as my Thesis would be related to Embedded Systems (not only Software, but full systems), the mother physical measures are also important. I received the instructions to choose a mother which can in some way "fit" in the robot; which can be physically supported by the Ronobova - I. Of course, I can implement something like a "backpack", so forth. But I cannot go beyond the "physical possibilities" of the humanoid. [This will make you laugh out loud...: as no student had picked up a project with 2 Intel Xeon Phi micros, I have suggested originally to create a centaur (to join the Robonova project with the Xeon Phi one)... whose body would be the Xeon Phi... Of course, I did not know the exactly measures of a Xeon Phi, but I wanted to test the teraflops intercommunication between the Robot (with one Xeon) with another external one..... The Thesis Director said: "the centaur is out of scope of an humanoid which has to walk just as a human..."
]
2. The second goal will be find another mother for video processing. Of course, Atmel 128 would not be possible.
I was suggested to implement one scenario like the following:
- Having a Raspberry Pi as core mother, which will provide the functionality of processing video.
- Also, regarding servos (RPi is not useful to control several servos), there is an external RPi shield capable to control many servos without overloading the RPi and just being commanded by 2 pins of the RPi). Other alternative, should be having the core RPi controlling a dedicated mother for servos like the 3024FX, the Arduino shield you have mentioned, Dagu Spider controller, etc., receiving commands from the RPi to process a direct control the Robot movements. Of course, if they can interact in proper way.
Well, as you can see, I am just "a beginner in the beginnings" but I am very happy with this project and hope to learn a lot from it!!
Any kind of suggestions will be very welcome!!
Thanks and regards.
AC.
Hi, VMatos.
Thanks for your email.
I am taking an Specialty on Embedded Systems, at University. My strongest skills are in C and C++, but at medium-high level programming.
Embedded Systems was always a pending subject for me. I am learning.
Well, my Thesis Director gave me a Robonova I, which is not working (seems some students damaged the motherboard some time ago).
So, I have 2 goals (as per my Thesis Director):
1. Having the Robonova walking again.
2. Having the Robonova processing images and "taking decisions" accordingly.
As I need to use University economic resources (this is a good practice to have students working in order way), I need to provide reasonable arguments before buying hardware, for instance.
1. So, for the first goal, the logical first step is to prove that the original motherboard is working / damaged.
Once proved, I have to suggest alternative motherboards (in case of damage) to replace it.
One alternative is buying the natural replacement for 3024 Mother, which is the 3024FX released by the same Korean manufacturer (HITEC provided the name of the real manufacturer to me; I contacted them, and they told me that 3024 mother and 3000 mother have been discontinued because they caused damages to the servos. Instead, they have release a FIX, the 3024FX, which is used by the Robonova II - Metal Fighter).
I appreciate your suggestion of Arduino-like for 24 servos and Dagu Spider also. Regarding Arduino, I suppose you are talking about some Arduino shield, which I did not know before your mention. I have also checked the Dagu Spider controller and it seems also a very good option.
BUT, look at this: as my Thesis would be related to Embedded Systems (not only Software, but full systems), the mother physical measures are also important. I received the instructions to choose a mother which can in some way "fit" in the robot; which can be physically supported by the Ronobova - I. Of course, I can implement something like a "backpack", so forth. But I cannot go beyond the "physical possibilities" of the humanoid. [This will make you laugh out loud...: as no student had picked up a project with 2 Intel Xeon Phi micros, I have suggested originally to create a centaur (to join the Robonova project with the Xeon Phi one)... whose body would be the Xeon Phi... Of course, I did not know the exactly measures of a Xeon Phi, but I wanted to test the teraflops intercommunication between the Robot (with one Xeon) with another external one..... The Thesis Director said: "the centaur is out of scope of an humanoid which has to walk just as a human..."
]
2. The second goal will be find another mother for video processing. Of course, Atmel 128 would not be possible.
I was suggested to implement one scenario like the following:
- Having a Raspberry Pi as core mother, which will provide the functionality of processing video.
- Also, regarding servos (RPi is not useful to control several servos), there is an external RPi shield capable to control many servos without overloading the RPi and just being commanded by 2 pins of the RPi). Other alternative, should be having the core RPi controlling a dedicated mother for servos like the 3024FX, the Arduino shield you have mentioned, Dagu Spider controller, etc., receiving commands from the RPi to process a direct control the Robot movements. Of course, if they can interact in proper way.
Well, as you can see, I am just "a beginner in the beginnings" but I am very happy with this project and hope to learn a lot from it!!
Any kind of suggestions will be very welcome!!
Thanks and regards.
AC.