SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio

Hitec HS-6635HB Servo Review

Ryan van Beurden puts the latest Hitec economy digital servo through a torture test and reports on how the servo took the abuse.

Splash

Introduction

The HS-6635HB is made with a rugged reinforced plastic case and exudes a feeling that it costs much more than it actually does.
The HS-6635HB is made with a rugged reinforced plastic case and exudes a feeling that it costs much more than it actually does.
Description:Standard Sized, high torque, digital servo
Motor Type:3 Pole
Bearing Type:Dual Ball Bearing
Gear Type:Karbonite (Reinforced Resin)
Size:1.6" x 0.8" x 1.5" (40mm x 20mm x 37mm)
Weight:1.83 oz (52g)
Torque (4.8v/6.0v):69/83 oz-in
Speed (4.8v/6.0v):0.18/0.15 sec.
Manufacturer:Hitec

Servos are one of the most important parts of the Radio Control "chain" between the stick under a pilot's thumb and the safety of the airplane he or she is flying. Over the years, I have used all sizes and shapes of servos from little sub-micro servos all the way up to big sailwinches that took 5 seconds to turn 180 degrees. I can't say that in all my years in the hobby that I have had no problems with the servos I used.

About 5 years ago, I got into flying high performance gliders and one of the first things that struck me was the sheer amount of money that sailplane pilots spent on their servos. It was not unrealistic to spend over $100CAD (about $80USD) per servo. I started flying a fully molded RnR Genesis model with some extremely cheap servos and I quickly learned that there actually was a difference between different brands and models of servos. The quickest things I learned were that servo centering is important, that a servo needs to be rugged, and that it needs to be extremely reliable.

Fast forward to today and I'm still at it in this wonderful hobby. I have been enjoying aerotowing and the latest plane in the fleet is an extremely overpowered Hangar 9 Ultra Stik Lite. There is a Brison 2.4 ci engine hanging off the nose of the heavily reinforced tug and I was very worried about how to safely control this plane in such a high vibration, brutal environment. With an engine that is 50% larger than the maximum recommended engine size, there had to be no compromises in the department of controlling this beast of an airplane.

About a half year before the tug project started, I read about a new type of gear that Hitec was putting in some of their servos. They were called Karbonite gears and were made of a reinforced plastic. They were advertised to have the low backlash of nylon gears while having the strength of metal gears. It seemed like a good concept, and with my background in mechanical engineering, I had no reason not to trust a well chosen plastic to perform as well as metal in a servo gear application.

I contacted Hitec to tell them about the towplane project and they quickly suggested their new HS-6635 HB servo. This servo is built in a standard size case and is sold as one of the lower cost digital servos in the Hitec range. I was interested in the cost of the servo and the seemingly good performance specs, seeing as I needed 6 of them for the flight surfaces on the tug. By keeping the cost down and the performance up, our goal of building an affordable tug would hopefully become a reality. So with that introduction, let's get started with torturing this servo…

The Servo

The Hitec HS-6635HB servo is built with dual ball bearings supporting the output shaft in a rugged-feeling, standard-sized, servo case. The servo feels very nice, with high quality 60 strand - 22 AWG twisted wire leads and it comes with an assortment of very rugged servo arms. When I first turned the output arm, I was absolutely blown away by the smoothness of the geartrain. It was silky-smooth, with no discernible backlash. So far, I was sold on the Karbonite gears. I took a good look at the specification sheet available here and then decided to get to the testing.

Testing

Since I have never done a servo review before, I decided to take some time to devise some good lab tests for the servo so that I could compare it to other servos in future articles. I wanted to get a quantifiable number for those things that the "pros" seem to always gush over. I wanted to prove things like good centering, powerful holding torque, and get an idea of the current drawn when the servo is pushed to the max.

I started by building a rugged frame to hold the servo securely on the edge of my workbench. Then I proceeded to use my Multiplex Profi 4000 transmitter and a Hitec SuperSlim receiver to drive the servo. I hooked up the receiver to a regulated lab-grade power supply so as to have exactly 4.8 volts (or 6.0 volts as the test dictated) at the servo regardless of the current drawn. I also hooked up a pair of Fluke digital multimeters to get an idea of the current drawn by the servo at the various phases of operation and to ensure that the voltage at the servo was remaining constant. Finally, I used a Mitutoyo dial indicator to be able to measure the repeatability of the servo horn position. This indicator has 0.001” resolution which proved to be more than adequate to see the variation in servos horn position.

Test Results

Centering Test - No load

This test involved operating the servo with no load. Using a switch input on my transmitter, I cycled the servo from center to one extreme and back to center 20 times and recorded the position of the servo arm after each cycle. It was impressive to see that under no load the servo continually came back to a position within 0.1 degrees of the median position, with only two trials out of 20 having a maximum deviation of 0.3 degrees. Even in this worst case scenario, the total error works out to less than +/-0.005” positional error on the servo horn hole furthest out from the servo pivot. That is pretty good repeatability/centering. Under no load, the servo drew between 4 and 5 mA.

Centering Test - 40 oz-in of torque

I took enough weight and hung it off the servo horn to cause the servo to put out 40 oz-in of torque. Note that by hanging the weight off a single arm, the worst case installation for the servo is being reproduced. This setup puts the maximum load on the output shaft and loads the servo case quite heavily. The only worse way to mount the servo is to have the short dimension of the servo case parallel with the applied load, but this is rarely seen in actual practice. In many cases, it's best to use push-pull linkages if possible, but I wanted to recreate the worst case (and realistically one of the most common) scenarios for servo loading.

With the application of 40 oz-in of torque, it is readily apparent that the servo does not reach the same position as it does under no load. At 4.8V, the servo had a median deflection from the median zero-load position of -4.7 degrees. While this may seem like a lot, it still represents only about 7.8% of the total servo travel, meaning that the control surface deflection is still a very large portion of what the pilot is asking for. At 4.8V under 40 oz-in of torque, the servo was drawing 222 mA.

It was interesting to note that at higher voltage, the performance of the servo was markedly improved. The angular deflection at 40 oz-in decreased to approximately -4.0 degrees and the servo only drew about 8% more current (239 mA). This makes a good case for using 5 cell receiver packs if you have the option and your servos are rated for it.

Centering Test - 58 oz-in of torque

As expected, as the load increased on the servo, the deflection from the zero-load center position increased. At 4.8V, the median deflection was -10.5 degrees and at 6V, it was -6.8 degrees. The currents drawn at 4.8V and 6V were 438mA and 460 mA respectively.

It is very apparent that as the servo is loaded more heavily, the use of 5 cells becomes even more valuable. While the current draw does increase at higher voltage, the price of extra current seems to be a small cost compared to the benefit of far better servo torque, speed and centering.

Centering Test - 77 oz-in of torque

At 77 oz-in of torque, it was apparent that the servo could no longer reasonably move the load at 4.8V drive voltage. While it still moved, it was probably only achieving about 50% of the motion that was requested of it. Luckily, Hitec has rated the 4.8V torque of the servo at 69 oz-in which is about where I would agree it ceases to operate in an acceptable manner.

At 6V however, the servo still quite happily pulled up the weight, achieving a deflection of -13.3 degrees from the zero-load position. It was sounding strained, but was still going strong. There was noticeable warmth in the servo, but at this point I had been beating on it for over 15 minutes, and the servo was still not what I would call dangerously warm. I was very impressed. While pilots seem to be obsessed with having the most torque in the world, I have to admit that seeing this standard sized servo pull up 4 massive nuts with confidence truly led me to believe that some people tend to err on the side of excessive torque that they feel they need in their aircraft. Based on the performance of this servo, I would not hesitate to use it in a good sized aerobatic plane (Hitec recommends the servo for aerobatic sport aircraft up to 78” in span). It will easily make a great aileron or flap servo on our tug and will be more than overkill on the rudder or elevator.

Absolute maximum ratings

One thing that impressed me about the HS-6635 HB was the massive holding torque. Even though the servo sounded strained to move the 77 oz-in load at 6.0 V, once it reached the steady state position, I could still give it a healthy twist with my hand before it totally let go of the intended position. This test shows the massive holding torque that digital servos have over the older analog style. At this fully loaded position, the servo drew 1280 mA at 4.8V and 1530 mA at 6.0 V. This indicates a need for a powerful receiver battery as several digital servos can draw down a reasonable amount of power in a short time if they are worked hard. In practice, the servos do not all draw this high current all at once and it is up to the end user to pick an RX battery capacity that is sufficient for their use. If in doubt, put too much battery in the plane. Our tug uses a 5 cell, 2400 mah sub-C nicad to drive all eight servos (7 of which are the HS-6635HB's). We will carefully monitor the RX battery voltage over the first several flying sessions to get an idea of the average current draw of our tug.

Finally, it should be noted that all the deflection values reported in this review are not simply the positional error in the servo, but include other factors like the small deflection of the servo mounting frame as well as the flex of the servo case, horn and output shaft. Though it is difficult to zero in on the positional error only, the tests done above are more applicable to the real world, since factors such as case and shaft deflection are deformations that will be seen in regular use.

Performance in the real world

Testing a servo on the bench is all fine and well, but it does not replicate many of the factors that a servo encounters in real flight. Things like the vibration of an engine and impact loading of ground strikes are hard to replicate. For this reason, there is still nothing like putting the servo through its paces in the air.

We installed the Hitec servos in the tug and flew it several times. The tow pilot, Adam Till, was extremely impressed with the performance of the servos on our heavily overpowered tug. Even under the extreme vibration and loads sustained in flight, never did Adam feel that the servos were not up to the task. They were noticeably better than the JR DS-811's that we were using before in the tug. The HS-6635's centered far better, while being significantly faster and more precise. Granted, the DS-811 is not direct competition for the Hitec HS-6635HB, but we were impressed by the improvement in performance that we gained from swapping out the servos. Based on the numerous flights we have had to date, we expect the Hitec HS-6635HB's to perform flawlessly for several years to come.

Conclusion

Hitec really has a winner on their hands here. The HS-6635HB is absolutely smooth, precise, and very powerful. There really is not anything more that can be asked of a servo that has a street value of a little over forty dollars. With its strong gears and quality construction, the HS-6635HB is a giant leap ahead of any standard servo. While it does not have the torque of some of the truly monstrous servos out there, it is perfectly suited to larger aerobatic airplanes in the 55-78" range as well as a perfect servo for large scale gliders. It is also ideally suited to any application where a more precise servo than a standard servo is required. I would not hesitate to use these servos in any of my planes and I have a sneaky feeling I will have a whole bunch of the HS-6635HB's in my fleet soon. From a dollars to performance perspective, they simply can't be beat.

Discussion

Reply
Thread Tools
Old Apr 23, 2005, 08:45 PM
Registered User
United States, CA, Burbank
Joined Aug 2004
79 Posts
Price

Great info, but no price. Did I miss it?

If these servos sell for $40 they are fantastic. If they sell for $65 they are just run-of the-mill. For $80 I'll go elsewhere.

Sorry guys, without a price this review is utterly useless. I rated it a 1.

Jim
jrf2 is offline Find More Posts by jrf2
Reply With Quote
Old Apr 25, 2005, 01:37 PM
Registered User
96S14's Avatar
Calgary, Alberta, Canada
Joined Oct 2002
739 Posts
Quote:
Originally Posted by jrf2
Great info, but no price. Did I miss it?
Thanks for your comments.

Perhaps I could have made the price a little more prominent, but it is stated in the conclusion:

"There really is not anything more that can be asked of a servo that has a street value of a little over forty dollars. With its strong gears and quality construction, the HS-6635HB is a giant leap ahead of any standard servo."

I will endeavour to put the price on the specs table in my future reviews.

Thanks for the vote.

Ryan
96S14 is offline Find More Posts by 96S14
Reply With Quote
Old Apr 25, 2005, 09:50 PM
Registered User
tailskid2's Avatar
USA, AZ, Phoenix
Joined May 2003
3,373 Posts
This was one excellent (IMHO) servo review! Congrats Ryan, ya did goooooooood!

Jerry
tailskid2 is offline Find More Posts by tailskid2
RCG Plus Member
Old Apr 28, 2005, 05:02 AM
Registered User
Joined Apr 2005
40 Posts
I got a lot out of the servo review, too. Are you planning on following it up with other reviews? It would seem to me that once you get everything set up, it should be simplicity itself to just swap in another servo. Of course, I'm sitting here behind my computer, so it's easy to make such brash statements.
kubark42 is offline Find More Posts by kubark42
Reply With Quote
Old Apr 28, 2005, 10:34 AM
Registered User
96S14's Avatar
Calgary, Alberta, Canada
Joined Oct 2002
739 Posts
Quote:
Originally Posted by kubark42
Are you planning on following it up with other reviews?

The idea was that I set up a format for making it easy to do further reviews on servos and for making comparisons in the future easy. As of right now, I'm not planning on doing any other pure servo reviews but I am talking to several vendors on some smaller servos (11mm - 13mm thick digital MG wing servos) As to whether that works out or not, time will tell.

Right now, I'm loaded up quite heavily on articles that I need to get written and planes that need to be built.

Thanks for the comments.

Ryan.
96S14 is offline Find More Posts by 96S14
Reply With Quote
Old May 05, 2005, 05:29 PM
Motorized lawn dart
Mopar92's Avatar
United States, MS, Columbia
Joined Nov 2003
2,152 Posts
Great review, however state the price next time. I really enjoyed it though. I found them online for $47 all day long... Thats a little rich yet for me.
Mopar92 is offline Find More Posts by Mopar92
Reply With Quote
Old May 05, 2005, 05:59 PM
Registered User
96S14's Avatar
Calgary, Alberta, Canada
Joined Oct 2002
739 Posts
Quote:
Originally Posted by Mopar92
Great review, however state the price next time.
This has already been brought to my attention, but thanks for the input. I did mention that they were a hair north of $40 in the conclusion. I feel that to some extent, giving a solid price doesn't make much sense with the volatility in Hitec prices. I'm sure the price I would have put in the article would have been obsolete by the time it went live on the website.

Here in Canada, the prices on Hitec are so competitive that companies are competing down to pennies and tossing in free shipping left, right and center. Hitec is extremely affordable stuff and a great value overall.

Also, here's a link to where you can get the servos in the US for less than $47:

http://www.servocity.com/html/hs-663...hi-torque.html

Hope that helps.

Ryan
96S14 is offline Find More Posts by 96S14
Reply With Quote
Old May 12, 2005, 06:07 PM
Official Boat Bum
Eddie P's Avatar
United States, NV, Reno
Joined Mar 2000
7,514 Posts
Quote:
Originally Posted by 96S14
Thanks for your comments.

Perhaps I could have made the price a little more prominent, but it is stated in the conclusion:

"There really is not anything more that can be asked of a servo that has a street value of a little over forty dollars. With its strong gears and quality construction, the HS-6635HB is a giant leap ahead of any standard servo."

I will endeavour to put the price on the specs table in my future reviews.

Thanks for the vote.

Ryan
Ryan, you are a gentleman for your reply. As someone who knows how much work can go into an article like this, a "1" is a joke of a rating for something like this. You get a "5" from me. Thanks for the info.
Eddie P is online now Find More Posts by Eddie P
RCG Plus Member
Old Jun 01, 2005, 09:37 PM
know it all
Miami, Fl
Joined Nov 2003
2,949 Posts
Great review!.. centering is critical in servos.

One quick question: do digital servos use a traditional feedback pot?.. or is it a encoder type device used instead?

Herm
hermperez is offline Find More Posts by hermperez
Reply With Quote
Old Jun 02, 2005, 11:49 AM
Registered User
96S14's Avatar
Calgary, Alberta, Canada
Joined Oct 2002
739 Posts
Quote:
Originally Posted by hermperez
do digital servos use a traditional feedback pot?.. or is it a encoder type device used instead?
Generally, it is still a precision pot that goes into an A-D converter. This means that digital servos can still be prone to positional drift with temperature, however as digi-servos trend to be more expensive in the first place, they tend to use better pots to begin with.

In my experience with optical encoders in heavy vibration environments through my line of work, I would not like to try one in a servo and expect it to be deadly-reliable.

Hopefully that helps.

Ryan
96S14 is offline Find More Posts by 96S14
Reply With Quote
Old Jun 02, 2005, 01:09 PM
know it all
Miami, Fl
Joined Nov 2003
2,949 Posts
Pots are not that durable either, they tend to wear around the center and you lose centering precision.. I even know some pattern flyers that will setup the servo with neutral off-center just to prolong servo life.. of course it used to be worse in the 70's, you had to replace the pots every few flights.. at least now they tend to last a flying season.

Anyways, that must be the reason they are still using pots instead of optical encoders. Thanks for the great review.

Herm

Quote:
Originally Posted by 96S14
In my experience with optical encoders in heavy vibration environments through my line of work, I would not like to try one in a servo and expect it to be deadly-reliable.

Ryan
hermperez is offline Find More Posts by hermperez
Reply With Quote
Old Jun 02, 2005, 01:11 PM
know it all
Miami, Fl
Joined Nov 2003
2,949 Posts
$50 is a very reasonable price for a precision servo.

Herm
hermperez is offline Find More Posts by hermperez
Reply With Quote
Old Jun 05, 2005, 03:36 AM
WORK!
hweird's Avatar
Joined Jul 2002
407 Posts
Ryan,

Great review. I like your scientific approach to testing!

What was the maximum deflection range during your twenty cycles? Did you swing the arm just to one side or did a complete cycle on both sides of center? How did you log the deflection points? Did you video tape the test? I am asking because I would like to test some of my servos and really like your methodology.

Did you suspend the loading weight from the servo arm with the same position of the unit as photographed? How would you measure servo speed?
Thank you,
Howard
hweird is offline Find More Posts by hweird
Reply With Quote
Old Jun 08, 2005, 10:20 AM
Registered User
96S14's Avatar
Calgary, Alberta, Canada
Joined Oct 2002
739 Posts
Hi Howard,

I used the switch output on my Mulitplex Profi 4000 and set it to go from -100% to +100%. So in essence, I was testing the precision of the servo's positioning at 100% travel. It can likely be safely assumed that the precision around center is the same though, since the position is simply dictated by the resistance that is fed back by the pot.

What the test entailed was to have the servo arm pointed straight to the side (parallel to the ground) with no weight on it. Then the switch was flipped and the servo lowered the weight somewhat past a vertical horn position. Then I flipped the switch again to have the servo hoist the weight and tried to see how close to horizontal it came again. So basically, a whole cycle was performed.

I logged the deflection just by hand. I didn't have a digital dial caliper, so I just fired the numbers into an excel spreadsheet on my PDA and then processed the data later on my PC. I did not video tape the test.

I don't quite understand your second last question, so if you can clarify a bit, I'd be happy to answer it.

Finally, servo speed I guess could be measured with a stopwatch and a microswitch if you really wanted to get a little fiddly. You could hook two switches up to the start/stop button of a cheap stopwatch and then have the servo arm click the switches on and off. It would take a little fiddling and you probably wouldn't be sure that you are measuring the time for 60 degrees of deflection, but if you were consistent in your switch positioning from one servo to another, you'd at least have an "in-house" benchmark to compare against.

The other option would be to videotape the test and then using some video editing software, you could look at the motion frame-by-frame and count the number of frames it takes for the deflection you are trying to measure. I think standard North American NTSC signals run at 29.97 frames-per-second, so you'd have a resolution of about 0.033 seconds from frame to frame...that may not be enough for what you are looking for, but it would be easy to try and may prove adequate for your needs.

Hopefully that helps. Sorry for the slowness of my reply. Let me know if you have any more questions.

Thanks,

Ryan
96S14 is offline Find More Posts by 96S14
Last edited by 96S14; Jun 08, 2005 at 10:26 AM.
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
For Sale Ebay: 4 x Hitec HS-6635HB Digital Servo, Airtronics 92778 8ch. Receiver Willard Aircraft - General - Radio Equipment (FS/W) 1 Aug 20, 2007 01:22 PM
FS: Hitec HS 55 servo Nik-rc Aircraft - Electric - Airplanes (FS/W) 10 Jun 18, 2003 04:49 PM
Hitec HS-81MG Servo - Stuck BlueFielder Radios 0 Jun 15, 2003 10:58 AM
WTB: Hitec HS-81 Servo's gromitvt10 Aircraft - Fuel - Airplanes (FS/W) 1 Aug 22, 2002 06:10 PM
Hitec HS-81 servo problem, But total success with Ken Hill's "Polly Pusher" dingo1000 Foamies (Kits) 1 Aug 09, 2002 11:50 PM