Hi All
I started designing my hexapod and i am working on a few fronts.
first, i built a “dumb” servo interface board based on a PIC, it controls all the servos and gets its input from a serial port that instruct it on the angle (PWM width) of each servo.
meanwhile, manage to understand the IK trigonometric mathematics and have the needed equations for the XYZ positioning calculations.
the “brain” is my computer that calculates all the needed angles and sequences and transmit it to the serial port.
the code is written in Csharp and hopefully, once complete, will run on a nano-itx board that will be located on the hexapot itself.
as for the mechanics, i started designing the hexapod using solidworks and encountered a few problems.
the robot will be built from a flat 6mm Plexiglas or ABS plastic and i havnt manage to get an idea on how to connect the coxa to the femur.
in my design, as can be seen from the pictures, i took the idea from the MSR-H1 hexapod concept but this concept seems a bit problematic.
would gladly take any recommendations on how to connect the coxa to the femur using the flat plexigalss/ABS plastic 2d parts.
You support the coxa servo between the two body plates, you have a connecting femur between the the knee and tibia servos. The Phoenix does this. I prefer to use yokes (c brackets) and a 2nd pivot (bearing) for each joint (aka SES).
but when you solved it, either by pairing ASB04 backets or some other way, the strain on the inner servo will be severe if you dont have a lower bodyplate … or maby you are planning to have such, just not being on the sketch …
… i am really looking forward to seeing a robot built round a nano-itx board !
Oh, I didn’t notice that there where no lower plate… highly recommended to add some sort of additional support (like a lower plate) or the hex won’t perform very stable.
I beleave he is asking for a way to add the coxa/femur servos together with a piece of plastic. He want’s tp cut the part from a flat plastic plate and can only cut it in 2D.
You can focus your design on Matt Denton’s MSR-H1 or do something like Zenta did on his original Phoenix. I beleave that in both designs you will need something stronger like aluminium to get it done. Or you need a pretty thick piece to get enough strength.
Thank you for your advices.
i have to say that for the last couple of months, looking at your projects (a-pod, phoenix code/robots) have been a true inspiration.
As for my hex -
well, i do plan on building a lower plate but removed it so i could clarify my problem when open this thread.
although i want to cut all the parts from a 2D flat surface, i will have to be flexible with my choices as i understand that there are some mechanical constrains.
all parts will be cut from a 6mm plastic or a plexiglass, which i am sure will give a good mechanical strength and rigidity.
as for the coxa part, i will try cutting it the way it described in my pictures (although i dont like this method) and if it will be a NO-GO, i will order the ASB-04 brackets.
still haven’t figured out what will be the affect (beside mathematical - angles and forces/moments) of a long coxa.
is there a recommended ratio for coxa/femur/tibia length?
i still need to select a nano-itx with a good processor so i will be able to run some more “complicated” applications on it (like face/voice recognition, synthetic audio out and some basic form of AI).
furthermore, although i believe that this will NOT be my last robot, i still want to design it properly (dimensions for all peripheral boards/batteries ect…) before start cutting it.
I am but a noob, but this is my thinking:
The Coxa is a bad thing, så keep it as close to zero as possible
Keep the femur and tibia allmost as long as each other so you can pull up your feet to the body whith the femur standing up at 90 degree. This will make your “horizontal working range” for your foot as long as possible
The length of femur and tibia is depending on the strength of your servos, at 100 mm femur a 10kg/cm servo only has 1kg liftingpower and that is quite enough but if you go up to 200mm femur whith the same servo you only have 0,5kg and it will make the walking unstable and the possibility to brake a servo will increse, so around 100mm for femur and tibia is a good choice (if you dont buy really expencive servos).
To long femur/tibia will allso make it difficult in some gaits as the legs will entangle eachother
And if you do want to do your own console, dont make them go arond the servoears, you need to be able to screw them to the console, sort of like this (i am not a good artist, sorry): http://nc.sverige.net/nc/robot/console.JPG
C brackets or ASB-04 brackets, even better, will keep the coxa length about as short as possible, as the two servos are basically back to back. Once you’ve got that, the femur and tibia aren’t that bad. A lower body plate, or a C bracket is necessary, as there is a lot of load on the coxa servo shaft if not.
A long coxa would multiply the load on the coxa servo shaft, and require a servo with more torque.
Hi All,
well, took some of your advice and (as originally planned) added the lower plate.
furthermore, i will add some spacers to the coxa mechanism for further durability and rigidity.
a for attaching the servos by its ears - i have some experience with such mounting technic (from my RC planes) and if done right, it will be a good solution.
i will try cutting these parts first and “feel” its robustness.
i will have to use some bearings for the lower plate for each leg to rotate freely, are there any suggestions on such bearings?
i had to enlarge the upper and lower plates to be able to accommodate a nano-itx board.
I think you have done the programming it alittle hard for yourself
Lets se if i can explain this
You have built the leg module all the same on both left and right side
Its easyer to mirror the front legs to each other, the middle legs to each other and the rear legs to each other
This way “go up” is the same direction on each servo, and “move leg to front” is the same
Makes the logic easyer
LM stocks the 3mm ID bearings.
I’d also add some spacers between the top and bottom chassis plates.
The yellow parts are your added coxa supports? They look like they would work.
By slot-mounting, you push the tabs (ears?) of the R/C servos into the slots of the coxa plates, and clamp the assembly together. That sounds do-able!
That just leaves you with the Tibia and Femur, which look similar to those of the Phoenix.
I would recommend working in some sort of rubber foot onto the Tibia, so that it doesn’t skid across a smooth surface. Maybe dip the tip into some sort of tool coating? My Tibias end in short rectangular section that conveniently accepts a rubber end-mill tube cap.
Hi PappaN - i agree, but its an easy fix as i can just flip each plate of the coxa structure of one side that will produce a mirror image of the other side.
(to be honest, i was easier in solidworks just to copy-paste each leg assembly and multiply it by 6)
Alan,
thanks for the advise, i will use a flange bearing with a small washer for as a spacer, of course the ling coxa spacers (the yellow ones) will have to be in the right length.
well,
started cutting the needed parts according to my solidworks design.
first part was a true disaster as i used an old blade on my scroll-saw… to much melting…
then, i replaced the blade with a fresh one and cutting was never easier.
i’ve extended the femur length and i hope i wont have ant troubles with the parts’ weight.
if needed, i will cut some of the material out for weight reduction.
i’ve attached the femur to the tibia using a round servo horn connected to the femur with four screws.
the securing screw that secure the horn to the servo is inserted through a large drill in the femur.
that way, everything is centered.
hey Raviv, what servos ur planning to use?
i see something what looks like MG995/996 or HXT12 -(i have 995) my opinion not best choice but price looks interesting.
how long your femur is?
) and what will be its affect ?
