KRS-4024HV Servo Vs. Bioloid AX-12+ What resolutions??

[NullARC]

I can't seem to find the resolution for the Kondo KRS-4024HV and the Dynamixel (bioloid) AX-12+. I know that I have seen them before.

What I mean by resolution is how many steps can be made while traveling it's full turning radius. The more steps, the 'smoother' the movement.

I know that the KRS-4024HV's resolution is better than the KRS-788HV servos used in the 2HV. I wanted to compare that to the AX-12's.

Thanks for your help!

- Marc

[NullARC]

In case anyone is interested, I have answered half of my question.

According to the article "Manoi Athlete: Humanoid Robots" by Lem Fugitt (known as 'tempusmaster' here), in the Winter '06 issue of Robot Magazine, the KRS-4024HV's resolution is 10 steps/degree over a range of 260 degrees. Opposed to 1 step/degree over a range of 180 degrees for the KRS-788HV. Quite a bit more resolution!

I haven't been able to locate this info for the AX-12+ yet.

[gdubb2]

Hi NullARC,

Try here:

http://www.tribotix.com/Products/Roboti ... X/AX12.htm

Gary

[NullARC]

Thanks Gary, that did the trick!

I had looked that page over and over and hadn't found the info I needed. When I saw your post I went there again, but this time I opened the PDF manual, and low and behold, there it was. DUH!

Now, it says that the AX-12+ has a resolution of 1024 steps over 300 degrees. But for some dumb reason I can't think of the mathematical equation for figuring out how many steps per degree that is. I need to go back to 9th grade algebra...

In Mr. Fugitt's article it says that the 4024HV has a resolution of 1800 steps over 260 degrees, while the 788HV has 180 steps over 180 degrees. Can someone who remembers the 9th grade give me the equation?

uuugghhh... the shame...

[gdubb2]

uuugghhh... the shame...

I would say so ....

Steps/degrees= steps per degree

[NullARC]

Yeah, that's what I thought. And that works for the 788HV servo. 180 steps divided by by 180 degrees is one step/degree. Which makes sense. However, when you divide 1800 steps by 260 degrees for the 2040HV you get not quite 7 step/degree. That is what has me confused. The article states that the 2024HV has 10 steps/degree. So that is why I thought I was doing something wrong. Glad to know I haven't forgotten that much! Any ideas as to why the article says 10 steps/degree and the math tells us about 7?

Also, I guess this solves the AX-12+ question. The step/degree ratio is about 3.5, give or take.

Ok, don't feel quite so dumb now! whew...

[heke]

I have a KHR-2HV, and I have noticed that the way it is programmed limits the movability more than the servos or their low steps per degree. I mean that it is difficult (or needs a lot of work) to program a smooth movement usgin H2H.

What I am missing with KHR-2HV is a tool to control the speed of the servo during its movement.

KHR-2HV truly can make smooth movements, but with H2H it needs a lot of steps. For testing I made some movements where the speed (or the position) of the servo follows value of sine wave. It looks nice, but needs 10-20 programming steps per 90 degrees or 180 degrees of movement. And all of these steps include information of all of the servos, which makes it difficult to make modifications on movements of other servos.

I wonder if the CPU of the KHR-2HV is the limitation (I mean that it is not possible to control individual servos separately) or is it just H2H.

[tempusmaster]

Any ideas as to why the article says 10 steps/degree and the math tells us about 7?

Sorry, I've been a little busy and hadn't noticed this thread.

The simple answer is that the H2H/RCB3 allows you to change the scaling. The figures you mentioned are the defaults as shipped and are good enough for most users, especially if they want to use the stock routines or exchange with other users. The figures quoted in the article are based on the Kondo RCB-3 pre-release seminar briefings and discussions with the designers. The numbers can go even higher, but it's not clear how useful or practical that would be.