PDA

View Full Version : [Project] designing high torque motor



robot maker
09-28-2008, 07:00 PM
looking at the HSR-5990TG that alot of us use for robotics and find out how to make one much better with half the current and about 5 times more torque
what HSR-5990TG has is 541 oz stall at 7.2 volts idle current 280 ma
my design is stall 2336 oz at 7.2 volts 150 ma
will post the design as i am finished it will be near the same size .771 x 1.5 x 2.5 high
cost will near the same HSR-5990TG is $125, mine about $100 or less

Enusi Malik
09-28-2008, 08:30 PM
Very intriguing, how far along are you?

4mem8
09-28-2008, 11:33 PM
I would love to see it Robot maker, What is your time frame with this servo?

lnxfergy
09-29-2008, 12:35 AM
So this servo will have 1/10th of the speed (~5s for 180degrees)? Or are you just lowering the idle current (in which case the stall current would be ~25A)?

Physics kicks in at some point.... A quick calculation shows the HSR-5990TG uses [email protected] at stall, or about 37W. 541oz-in = 33.8ft-in = 3.8nm, 0.17s/60degrees = ~1RPM = 3.14rad/s, 3.8nm*3.14rad/s = 24W = about 65% efficiency. So um, its not like the HSR5990 is really inefficient... something else has to give...

-Fergs

Adrenalynn
09-29-2008, 01:31 AM
My only regret is that I can't give you more than 1x+Rep for that post, Fergs.

Alex
09-29-2008, 11:21 AM
+Rep from another mod;)

DresnerRobotics
09-29-2008, 11:45 AM
Personally I'd wait until I had the proof of concept built with pictures/video and verifiable proof of how much torque and current draw the unit can produce/draw before I posted ;)

Adrenalynn
09-29-2008, 11:56 AM
You're such a big meanie, Tyberius!

As long as you have all the numbers and science to back it up, a theoretical post can be intriguing, imho. "Thinking Man's Design" and all.

Thanks, Alex! I really felt compelled to up-rate that post twice - thanks for making that a reality!

Alex
09-29-2008, 12:15 PM
You're such a big meanie, Tyberius!

Yeah, I know right? I think I'm going to have to put some tacks on his chair or something... maybe nails?

DresnerRobotics
09-29-2008, 12:23 PM
As long as you have all the numbers and science to back it up

Exactly.

robot maker
09-29-2008, 12:27 PM
So this servo will have 1/10th of the speed (~5s for 180degrees)? Or are you just lowering the idle current (in which case the stall current would be ~25A)?

Physics kicks in at some point.... A quick calculation shows the HSR-5990TG uses [email protected] at stall, or about 37W. 541oz-in = 33.8ft-in = 3.8nm, 0.17s/60degrees = ~1RPM = 3.14rad/s, 3.8nm*3.14rad/s = 24W = about 65% efficiency. So um, its not like the HSR5990 is really inefficient... something else has to give...

-Fergs
not using the same motor for my design,so the ouput speed wont be the same ,it will be 6rpm at no load speed so about .1 sec at 6 volts
idle current is very high for hsr-5990 380 ma,they dont say how much at full working efficiency
like the motor i am using witch is 500 ma and idle is 180,so i am thinking hsr-5990 will about 1 amp


on time frame it will be in few weeks,since still building the johnny five design and will be using it for the arms, i ordered the gears,pins,bearings and rest of the parts except for alumium plates
it will be a little higher then stated because of the pot i have doesnt have a longer shaft,could use one with a longer shaft but will be much longer in width,need a .5 round pot with atleast 1 1/2 shaft length and ss or aluminuim,found many at 1 1/2 round with 2" shaft

yes that can be done fairly easy,ill be posting as soon as i make it,made many in the past at my work

tybererius said

Personally I'd wait until I had the proof of concept built with pictures/video and verifiable proof of how much torque and current draw the unit can produce/draw before I posted

lnxfergy
09-29-2008, 12:36 PM
6rpm at no load speed so about .1 sec at 6 volts
What? .1 sec for what? If the final output is 6rpm, that is about 1/10th the speed of the HSR-5590. Thats a full 2 seconds for 60 degrees of rotation.... really freakin slow... great though for building a robot that can wave at the audience like the queen of england...




idle current is very high for hsr-5990 380 ma,they dont say how much at full working efficiency
like the motor i am using witch is 500 ma and idle is 180,so i am thinking hsr-5990 will about 1 amp


An amp? Try 5Amps at stall for the HSR-5990. Also, I would think most builders barely care about idle current - its the holding current you are dealing with most times (especially with arms)- which is way higher....

-Fergs

robot maker
09-29-2008, 12:45 PM
so hsr-5990 will draw 5 amps ,battery you are using will be dead real soon ,time 6 of them on a arm that most are using

motor i am using at 1.9 amp stall ,compared to 5.1 am stall on hsr-5990 and stall torque is 534 oz compared to 416 oz,so just using the motor by its self is much better then hsr-5990
might have it wrong for 6prm
mostly looking higher torque at lower current



What? .1 sec for what? If the final output is 6rpm, that is about 1/10th the speed of the HSR-5590. Thats a full 2 seconds for 60 degrees of rotation.... really freakin slow... great though for building a robot that can wave at the audience like the queen of england...




An amp? Try 5Amps at stall for the HSR-5990. Also, I would think most builders barely care about idle current - its the holding current you are dealing with most times (especially with arms)- which is way higher....

-Fergs

also did some comparing of speed of the motor i am using without phototachmeter and both motors are near the same speed
looking to buy a phototachmeter since it will be hard to use the equipment at work, but my design is using near the same idea that servocity has on thiers power gearboxes using hsr5990 and 5-1 ratio with feedback pot ,without any tests yes seem i should have near the same speed ,maybe more/maybe less .75 sec,dont know how i got 1sec and will have more then 5 times as much torque
with using half the stall current
i guess a easy way to test my design since i do have 5-1 gear box with hr-5990tg is to remove the servo motor and make a simple plate for the motor i am using

robot maker
10-01-2008, 08:57 PM
sometime this weekend will have photos of my motor attached to a servocity robotzone servo box 5-1 gear box for testing the torque,since it will be about the same as my design 4-1 ,but alot smaller
speed about the same maybe a little slower torque alot more over 2600 oz,it will have weight on it to test the design and using alot less current too stall current on mine design is 1.9 amps on hitec alot higher 4.2 amps at 6 volts lock/stall,after the test will have photos of my design

4mem8
10-01-2008, 11:18 PM
Cool, It will be interesting to see the specs and how they stand up to tests.

robot maker
10-01-2008, 11:47 PM
will be testing for current verses torque,how much current at the given torque,i am thinking of using a 12 inch arm if i have one and add a wire to hold the weight,if i really had the time i would make a better design to hold the weight ,like a basket,maybe with the final design and measure the speed,i dont care to much for speed on my johnny five,mostly saving battery power and having more torque
if speed is too slow can always go to 2-1 ratio and at the same time have more then double the torque at half the current

also the final design will have the same plate for mounting servo as the hsr-5990tg