I am programming for a Lynx robot arm containing the following servos:
HS 645,HS 475, HS 422
If I understand correctly, each servo type has angular limits of travel enforced by an obstruction on the plastic mechanism, and these limits of travel correspond to pulse widths sent out by the servo controller. I therefore assumed that I would be able to find specifications for each of these servos of the form: (just making up numbers here) maximum travel: ±70degrees;-70degrees= 750 microseconds; +70degrees= 2250 microseconds. I have been unable to find such specifications. In particular, on the Hitec site I can’t find these particular servos listed and, for the servos that are listed, I don’t see where the travel limits are specified.
Where do I find this crucial information?
Also, are these servos all analog servos? How would I know in general?
The servos you mentioned are analog. All of them will do +/- 90°. The servos will be centered with pulses in the neighborhood of 1500uS. The servos will be around the extreme CW position at around 2250uS, and extreme CCW position at around 750uS. As the servos are analog the exact positions need to be found by experimentation. For example I grabbed a 422 servo and tested. At 1480uS it was centered. At 550uS it is at 90° CCW, and at 2400uS it’s at 90° CW, in reference to centered. But another one off the shelf may be slightly different. These are the facts.
.Jim,
Thanks for the response. That takes care of my immediate needs, but I’m still confused about the general situation.
Suppose I was designing a robot and a particular joint of the robot required a servo that went from -40degrees to + 80degrees. In usual engineering practice, one would get out a catalog or look online and find a device with the appropriate range, torque , acceptable price, etc. How could I do this? I can’t get a sample of every available servo and measure the range!
Also , when you say that the range of a servo type will vary from sample to sample, how much of a variation are you talking about? A few percent? Do you mean that the available angle range actually varies or do you mean that the corresponding microseconds limits vary?
Regards,
Joe
to be honest it sounds like you are making the classic error of mistaking hobby servos for something you would use in practice as an engineer. they are not. they are largely consumer level devices that are vaugely required to conform to a general standard of +/-45 degrees over a pulse range of 1 to 2 ms with a 20ms frame, 1.5ms being the 0 degree point of reference. the fact that through experimentation we can determine they can do +/-90 degrees over a 0.5 to 2.5ms pulse range is all well and fine but you can just as easily find some that will only do 178 degrees or the 1.5ms isn’t the center of a +/-90 degree range because the servo is only specified to be +/- 45 degrees. if an analog servo is specified by the manufacturer to be +/- 90 degrees then it is a sure bet they will do something more just so they do not have to tune or calibrate or even characterize them when they are manufacturered.
Consider that even an expensive hi-torque digital hobby servo would be a tremendous value at $100 if it came with torque curves and precise specifications that a manufacturer could possibly be taken to task with for failing to meet the specified performance levels. It just isn’t going to happen. you can take the servos and make measurements and come up with calibration data and put your own dollar value into them but the problem is the next batch you buy might vary by 10-15% and still meet all the specs the manufacturer has given. that is the point, they are specified in such a way as to protect the manufacturer. you however are getting an inexpensive, relatively speaking, product that is very generically specified such that choices between different models can be made. if you want better you need to take out your checkbook and go shopping.
All standard size analog servos from Hitec do 180° rotation. If you need less range, then no problem, just limit the range in software. If you need more than 180° then you must resort to gearing. This is not easy to do, and the servo’s torue will be reduced. Torque and price for these servos is published and it’s easy to figure out.
Digital Standard Size Servos (only capable of ~140° out of the box, but can be programmed for 180° with the HFP-10 servo programmer, $154.99)
HS-5475HB - 76oz.in. - $34.99
HS-5645MG - 168oz.in. - $54.99
HS-5955TG - 333oz.in. - $114.99
Digital Standard Size Robot Servos (capable of 180° out of the box, and the HMI Servo Programmer can alter the parameters, $24.95)
HSR-5980SG - 180° - 417oz.in. - $109.99
HSR-5990TG - 180° - 417oz.in. - $124.99
Digital Robonova Robot Servos (capable of 180° out of the box, and the HMI Servo Programmer can alter the parameters, $24.95)
HSR-8498HB - 180° - 103oz.in. - $59.95
HSR-5498SG - 180° - 188oz.in. - $69.99
The variation is only a few degrees. The main contributor to the variation is with the potentiometer used for feedback. They are used in all servos, analog or digital, and the servos are at the mercy of the accuracy of the pot manufacturer.
Eddie and Jim:
Thanks to both of you for the detailed clarifications. Especially Jim for putting together all those numbers. I am a happy camper now.
Regards,
Joe
When I originally posted this information I suspected as we all did that the potentiometer was the weak link in servo response curves comparing one unit to the next. Well I finally have indisputable proof of this. I modified 6 HS-422 servos so they could be used as input devices. The modification was straight forward. Open the servo, remove the PC board, short the motor wires together, and finally solder the servo cable directly to the pot. Simply connect the pot green, yellow, red to the servo wire black, yellow, red. Now for fun I connected them each to the Bot Board and Atom Pro processor and did a simple read of the A to D port while adjusting from end stop to end stop. I got the following results.
Sure there is bound to be some other sources of errors in the driver board, which is now surface mount on the 422 btw, but it’s very clear that the pot is a major contributor of error comparing one servo to another.
Sort of related, but I have a question about servo “hold”. I have a couple each of HiTec HS-311 and Futaba S315 (bought used in a hobby consignment shop) servos. I hooked them up to
a Parallax servo controller board and the HiTec, after moving into a position would “hold” that position…I try to move it and it hums and resists. The Futabas on the other hand will just be
done after the move. Once the position command completes, I can freely rotate them.
Is this something unique to HiTec (the holding) or are the Futabas just old tech?
Also, is there a difference in this respect (the “hold”) between analog and digital servos?
I’m in the process of determining which servos I want to buy to use in robotics and am trying different types. I’m sort of starting on the cheap and need this info before I go
buy a shoot-load of servos.
There must be something wrong with the futaba servos. All servos are supposed to hold position. A servo isn’t much good if it can’t hold the position it was commanded to.
The difference between analog and digital holding power is significant. The analog version updates the motor voltage pulses every 20mS, digital servos update the motors pulses at a much higher rate. Probably around 5khz, or every .2mS.
I’m looking to buy a bunch of Hextronik 5010 servos - some cheapie brand at Hobby King that seems to get good reviews - only 5.95 each! Dual BB, Analogs and
have 6.9Kg torque. For the price (and as long as I can get them to turn 180 Deg.) and they HOLD, they’ll be a perfect start for my experiments (once I get all the
brackets, processor, etc. from here).
The servos are physically too tall to fit the brackets. They use an oversized motor to achieve the torque they claim. There are a lot of posts here concerning the quality, or lack there of.
Huh. My bad - lacking actual measurements in your bracket description (other that ‘Fits standard size servos’) I guess I assumed this was a standard size servo.
Also wasn’t on the “beware” list (Bluebird or GWS) warning about oversized.
So “standard size” (as relates to your brackets anyways) would be HiTec?
We do offer the 3D models of the brackets, but I’m not expecting everyone DL them to check for fit. The brackets were designed to fit Hitec servos. Futaba servos will also fit. There are a couple other manufacturer that may fit, but I haven’t tried them. We will update the beware list, but the truth is all of the cheap servos are made oversized (compared to Hitec or Futaba) and will not fit the brackets.
Hitec and Futaba are considered the standard in RC. The first two sentences makes it pretty clear when we say standard size, we are referring to Hitec.
I will have it reworded to eliminate any ambiguity. I will simply list the servos they are known to work with and a list of servos that are known not to fit. I’m truly sorry for any inconvenience this may have caused.
