Slightly off-topic, how are Kondos and Robonovas getting away with weaker servos? It has to be the power(torque since we have rotating joints)-to-weight ratio right? I’ve often wondered what the performance look like since many competition bipeds that I often see from Tempusmaster’s videos are one with more powerful replacement servos (even the Robonova’s $69.99, 200+ oz-in servo will make the bot quite formiddable).
It would be VERY interesting if we can get a comparison of weight sans servos between biped robots out there. Comparison of weight for just the frames and brackets for the same number-of-DOF class. For instance, the weight of just he frames (with electronics but no battery since that can vary from LiPo to NiMH and cell count and size) for:
SES-based humanoid with Lexan chassis (biped pete!)
SES-based humanoid with no Lexan chassis (using purely SES brackets for chest like the waist rotate biped)
Kondo KHR-2HV (the 1HV have two extra DOFs which would be unfair)
As you suggest weight is the biggest factor. They are small and lightly built. Of the smaller humanoids, the Futaba seems to get around the best from what I’ve seen.
I don’t have the weight for our bipeds. I can tell you our brackets are made from 0.06" material. The robonova brackets are made from 0.046" material. We never intended our SES based Bipeds to be as light as possible, but rather went the strong as hell direction instead. Plus our bipeds are simply bigger! Really an apples / oranges thing here.
I can’t agree more. LM’s brackets are fantastic! They are very reliable and strong. The all ses-based bipeds are very formidable in the biped humanoid world. It’s no coincidence that many of the competition-grade bipeds are mostly aluminum-based, as it can withstand punishment and abuse…
I agree with tom, these ses brackets are amazing. I hope to make my humanoid out of them. I also really think it would be cool if there was a bracket for two servos on the same plane like on the bioloid hip/ankle.
thanks for all the input guys. for anyone reading this DONT BUY CHEAP SERVOS FOR BIPED PETE! it will just cost you more in the long run.
looks like i will have to slowly build up my servo collection. 5645’s seem the go for me. once i get Pete outfitted with some decent hardware i’ll be back with more questions im sure
Here is a height comparison of the robonova, Biped Pete, Biped BRAT and Lynx209… You can see even the BRAT is almost as tall as the robonova. I don’t currently have a Biped Nick (waist rotate) assembled for a height measurement.
Never realized the biped scouts were almost as tall as the humanoids. God I miss the 6DOF leg of the scout, it was so articulate… Gotta save up for a HSR5990TGs and restore my biped scout back to its original glory… Too many bots!!! They are like kids, each one needs attention!
while programming pete to walk, i have been using values that i have previously defined such as angle_45, angle_90 etc to move the servos from their “straight” position.
so here comes the question. from the center position is the value to move say, 45 deg, different for left and right? it seems that in one direction the servo moves exactly 45 deg, but in the other direction it is not so precise. whats up with that???
You’ll find lots of threads on here saying the same thing. There is no exact center, the range isn’t exactly 180 degrees (it’s a bit more), and there is usually a little more range to one side than another. Center it as best you can, then for your end points choose a pulse that gets you the same distance from center on each side before it starts binding on either side. Sorry, that’s the best you’re going to do with cheap hobby servos.
i’m not too concerned with where the actual servo center is. i have set values that i deemed to be my “center”. they are not the actual servo center.
for example:
[code]leg_straight con 4000
degree_45 con 5000
hservo[p1\leg_straight+degree_45\250]
hservo[p1\leg_straight-degree_45\250][/code]
these two commands do not give ± 45 deg from “center”. any idea why? could it have something to do with the fact that in the second hservo command i have ventured from positive numbers to negative?
I assume you are still wanting to use Atom Pro and Hservo. If you look at the new BRAT tutorial you will see a program James wrote that clearly illustrates how to set up servo offsets for Hservo. It’s really worth the time to do this! If you take the time to center the servos in software if you ever need to replace a servo it’s a simple matter of calculating the offset and changing one value in all of your programs. Don’t get bogged down with the range being different in one direction compared to the other. The difference is minor and not worth the trouble trying to correct it. But setting up the offsets is a BIG improvement!
thanks for the link. i like the way some of the code James wrote has been setup. i was looking for a way to easily incorporate the speed calculation for the servo movement into my code. the brat code looks like a winner:)
On my Brat, I used the remote control to adjust the center points. If you hit the menu button on the TV remote, it went into a mode, pressing number keys selected which servo (1,4,7 on one leg, 3,6,9 on the other) and I used the Up/down vol and channel to adjust. Pressing enter or return stored the new value into the EEPROM of the Atom Pro, which the program reads out at startup time. The code worked reasonably well…
i just received my new leg servos i was hoping to install this weekend. due to a mix up i ended up with 2 HS-5955TG’s instead of all HS-5645’s that i was going to use for every joint. since these are double the torque, what leg joint would you guys recommend for the 2 higher torque servos?
I didn’t want to open a new thread to ask this and I’ve gotten some good info above as well. I am about to order some parts to upgrade my BRAT into a Biped Pete - well, legs and torso only at this point. In Step 5 of the assembly guide for the torso lynxmotion.com/images/html/build087.htm, are those ASB-09 “C” brackets? I wasn’t sure since the Multi-Purpose servo bracket doesn’t seem to be needed.
I finally have my Pete (with no arms - just shoulder servos with horns) together and working through Lynxterm and VB 2008 over BlueSMiRF with a 5 cell pack as the head . He is not walking well enough to take video yet but all servos are responding and centered pretty much now. I am using all 645’s and they are strong enough but I don’t have the weight of the arms yet and I’m not trying to do acrobatics either. I used those crap TowerPro 995’s for the shoulders and neck and I may keep them there as I build the arms but I will not order another one as the horns don’t fit the “C” brackets. I’ll probably gradually replace all the servos with 5990’s but it’s not in the budget yet. Well, here’s a pic of it currently. Thanks again for all the advice out here. I’m so glad I went the LM route with the biped as I’ve been happy with their rovers and polypods in the past!
Plus our bipeds are simply bigger! Really an apples / oranges thing here.
. He is not walking well enough to take video yet but all servos are responding and centered pretty much now. I am using all 645’s and they are strong enough but I don’t have the weight of the arms yet and I’m not trying to do acrobatics either. I used those crap TowerPro 995’s for the shoulders and neck and I may keep them there as I build the arms but I will not order another one as the horns don’t fit the “C” brackets. I’ll probably gradually replace all the servos with 5990’s but it’s not in the budget yet. Well, here’s a pic of it currently. Thanks again for all the advice out here. I’m so glad I went the LM route with the biped as I’ve been happy with their rovers and polypods in the past!