I have a project that needs a lot of torque and precision. I tried pic-related d954sw which is the top of hobbyist servos and it's not doing the job. Specifically, speed on way down is way faster than speed on the way up (at least it gets up unlike my dick). It also bounces past target angle when it does manage to speed up. In other words I am 100% sure it can't handle the weight. I also tried "70kg" Chinese servos and they have less power than this "30kg" servo.There are $300 ClearPath servos that will do the job, but I can't go up in price / servo. My question, to those who are "in the know" is this: Am I paying for the small form factor? In other words, are there "cheap" (<$100) servos that will do the job if I am willing for it to be bulky and supply more power? If yes, could you find me an example that will be over 100kg/cm of torque? If not, what are my options?
just dont use a servo you. use a dc of some sort with a gearbox.
>>2885537That seems like a lot of work - all the big boys are using fuck-hueg servos too. Big Dawg just uses servos. They are $500 each though.
oh help?desu im in the middle of reinstalling linux for the 100th time cause im being hacked and windows wont install. i got my stuff backed in the cloud though so yeah ill install ubuntu with selinux or app armor and have to use freecad i guess.
>>2885544THICC. Try Fusion 360 unless you are very used to CAD.
>>2885529Learn to use stepper motors. You can buy steppers with closed loop control for $30-ish on Aliexpress, then even program the onboard chip to do whatever you want it to do. If you aren't capable of learning, then $300 servo it is.
>>2885548How about something more specific? In my experience steppers pretty much have the same cost/power as servos. I can program and learn just fine, but this isn't even close enough as a lead to start on.
>>2885540the formula for the hip motor is torque(nm)=weight of the robot in kg*height of the leg in meters*sin(45)45 is the degree of motionyou can get a worm geared motor with the torque or make the gears yourself. but desu you should just get a premade from aliexpress.
>>2885547no im serious im beign hacked i need windows for fusiin 360
>>2885562Ah, got you, sorry man.
>>2885552>pretty much have the same cost/power as servosWell you don't have much experience then. Any basic bitch $30 planetary geared NEMA17 can do 10Nm, let alone NEMA23 or worm gears.
>>2885568I will look into those now, ty.
wtf even is this thread
>>2885568I am not seeing anything that indicates NEMA23 steppers have better torque than tiny-ass kiddy servos.They also have 1.8 degree / step accuracy vs .1 degree / step on servos.> Any basic bitch $30 planetary geared NEMA17 can do 10NmI am seeing 2Nm, maybe, which is worse than standard-size servos. That's for $80+Will program robot-waifu to suck you off if you can find a stepper that will do 10nM or 1500oz/in. under $150, which was the question for the servos.
>>2885706https://www.amazon.com/STEPPERONLINE-Digital-Stepper-1-8-5-6A-20-50VDC/dp/B074TBMC7Nplushttps://www.amazon.com/Controller-PEMENOL-Micro-Step-Resolution-Adjustable/dp/B0BK1GYBR1drivinghttps://www.amazon.com/Torque-Stepper-Motor-425oz-Router/dp/B00PNEPW4Ccombined withhttps://www.amazon.com/STEPPERONLINE-Planetary-Gearbox-Gear-Ratio/dp/B0BPGLLW11425*5=2125 oz/in but derated for chinkshit is probably 1500obviously you dont understand microstepping so look that up
being a servo has nothing to do with motors power output nor does nema rating. this thread is but copium.
>>2885712Thanks anon, at least you make some sense. > obviously you dont understand microstepping so look that upI really don't, I will google it. I assume that's just a way to fine-tune control the stepper?At the end of the day, is the transmission (let's say the 1:5) basically:> 1/5th speed> 5x torque> 5x precisionor am I missing something there? > increasing torque, reducing speed, accurately positioning and controlling reproducible machinery.Looks like their notes pretty much align.Anyway, the last link is really the eye opener for me, I thought I would need to do gears and a belt like a KEK.Thanks again, learned a ton. I kind of need the speed too, but at least this is something.
>>2885529https://www.pololu.com/product/4847Actual 30kgcm (there’s also a 378:1 that does 90kgcm). Encoder does 6500 pulses per rev on the output shaft. You’d need to write your small PID control loop on an arduino or something and it should lift easily. But that’s a 5 minute job and allows you to compensate for gravity
>>2885756>5 minute jobill do your mom in 5 minutes
>>2885529Have you tried getting exercise, fresh air and some sun? Also a water pump goes a long way into restoring function before looking at a cialis and/or low dose of test E
>>2885529How fast do you need it to actuate? There are a number of solutions here:10:1 reduction gear or even more, you can 3d print one (fdm is stronger) like a planetary system.Chunky stepper motor and assume it gets into desired position + reduction gear if needed Hybrid servo sensor on stepper motor + gear as aboveUse a brushless motor with an encoder for position feed back and massive reduction ratio.. or even just a DC motor, pi Pico etc, shit loads of torque if you only need to complete 1 revolution in 2 seconds etc
>>2885774> How fast do you need it to actuate?I really wanted 360 in one second, which is what I get with the d954sw servo under light load, but the more I think about it the more I think 1 second for 90 degrees will do, so if I got 360/5 per sec with a 1:5 gearbox, I would probably be fine. It kind of ruins the idea a bit but might be passable at that speed. Ideally, I would be over 180* a sec.>>2885756Thanks anon, that's actually a good price point - 1/2 to 1/3rd of what an equivalent servo goes for and opens up budget for other things potentially. I am using an arduino already and I don't care if I have to code a lot. I wrote some servo-easing code and realized its pretty much better than the servo easing library the arduino recommends.>>2885771Oh, I forgot I made that sad excuse for a joke in the OP. For anyone reading, basically the meta seems to be not to have excess weight, even 10-20 lbs will fuck with hormones. Also cut porn and never touch your pp before it's fully hard. Then good diet and lifting as secondary steps. Obviously running for cardio doesn't hurt either, but that's more if you lose erections. tldr is too much fat and too much porn for 99% of the people having problems.
>>2885774> 10:1 reduction gear or even more, you can 3d print one (fdm is stronger) like a planetary system.Gear slop! LMFAO.
>>2885845Not the guy you are replying to, but I am looking at these now: https://www.servocity.com/servo-gearboxes/Specifically the Stingray-2. Pretty fucking slow, but also minimal changes on my side. I do wonder what would happen if I just had a friend print / CNC some gears out of plastic.
>>2885552>steppers pretty much have the same cost/power as servosServo is a canned application. If you don't like the response characteristics of a servo, you're stuck. Stepper controllers (good ones) let you fiddle around with PID tuning.
>>2885864>Servo is a canned applicationmaybe for these goofy little geared toy things. an actual servo is nearly infinitely tunable and orders of magnitude better than a stepper
>>2885871>ACThis is a DC neighborhood. Fuck outta here.>>2885864I don't really understand desu. I thought they had x number of positions and that was that. Or you mean for speed and torque?In either case, my main take away from this thread is that stepper motors seem like a cool complex term to throw out but dont really offer anything more in terms of performance. Though, it does seem they have been around longer so there are slightly better commercial offerings.Also this guy is my new hero >>2885712 he can fuck my mom if he wants.
>>2885864>>2885880with all your drivel i dont think you know what ac, dc, steppers or servos are. i wouldnt be surprised if youre same fagging the whole thread like cock gobbling larper.
>>2885880youre okay with spending $70 on a single stepper motor?oh well atleast youre a productive consumer
>>2885806A better question, does your precision mean ‘stop at position X precisely’? If x is always the same one/two positions, then just get any DC motor and four switches. First switch changes to low speed, second switch is the end position. Put them on an rail to account or whatever to adjust for slack and stopping time. We stop 2t+ lifts and cranes at 5mm precision this way
>>2885529Had a similar issue, not sure if its exactly the same as what you're describing but I think it is. It stems from the fact that the control scheme build into those cheap pwm servos aren't really sophisticated enough to handle much other than "eventually get to the setpoint, maybe". You can brute force your way past this by just getting an overpowered servo, but another option if you're inclined to a little manual work is to retrofit the servo with better closed-loop control.Take a look at https://www.youtube.com/watch?v=_4mrb2T706s - if that same kind of movement looks like the problem you're having, then https://github.com/adamb314/ServoProject/tree/main might be worth your time
>>2886276This is really cool, thanks. Looks like the full explanation is here https://www.youtube.com/watch?v=SioCwvR_PYY and it's quite a bit of work. Furthermore, something I should have shared earlier (but was implied by "not enough power") is that the servo is outright faster on the way down than up, which tells me it's really a power issue. I do want to brute force it though, since following the vid is a REALLY involved project. Still, learned something new, ty!>>2886232Unfortunately I need precision over the whole servo range, it's more of a problem like in the video above where I need to point to certain, random angles, in random order.>>2886226I am waiting for your 1500+ oz-in, .4 of a degree or better, 120 degrees in a second or better solution for under $150 (I know you said $70). Enlighten us.>>2886160Here is your (you).
Working on a home project, tiny wee 3-axes home cnc for hobbyist jobs, Aircraft frames and the like - from low density woods, foams and ideally acrylic, Having trouble with the spindle motor and power supply. Ideally this would be as dirt cheap as possible. looking for options as to what I can do. Do I make the power supply? would it be worth trying to make my own motor?Less than half a meter on all sides. Aluminium extrusion, lead screws, nema17 motors, arduino uno 3 w/t drivers. i am a tight fisted bastard so act accordingly
>>2886518trim router
>>2885880>I don't really understand desu. I thought they had x number of positions and that was that. Or you mean for speed and torque?a servo is just a motor with feedbackyes typically hobby servo has set angle of travel because that is what their application requires and is used togenerally you can call anything a servo if it has some kind of feedback where you 'ask' it to do something specific like a certain speed or a certain position. you can call a stepper + controller a servo in fact thats what many industrial servos are.
OP, there's a third option no one's mentioned (and admittedley given your use case you might not have the room to implement it); a regular DC motor (geared or not) with an optical encoder wheel. You'd still have to do some programming to determine position, but the hardware end is reasonably easy to set up; you can even salvage most of it for basically free (encoder systems are usually used in desktop printers for the paper feed positioning).
>>2886547> set angle of travel because that is what their application requiresIt’s mostly because using a potentiometer is the cheapest feedback and easy to integrate> stepper + controller a servo in fact thats what many industrial servos are.In what industry? All I see in food/beverage/packing machines is permanent magnet synchronous motors with encoders for servos
