By continuing to use this site, you agree to our use of cookies. Find out more

Testing clone of the TowerPro MG996R Servo

Various mechanical tests and inspection of the servo

All Topics | Latest Posts

Search for:  in Thread Title in  
John Wh27/10/2016 00:39:15
avatar
12 forum posts
7 photos

I was asked to test these for an Amazon review. Thought I'd share my findings with you. The main thing to note is that some adverts say that these have double ball bearings- they do NOT!! They only have a bushing that does not stop a lot of shaft wobble.


In Summary:

Maximum torque measured = 10 kg.cm @ 6.5V
Speed: Slightly slower than a Futaba S3003
Freeplay = 0.7° rotation and 1.4° on the shaft
Bearing: One brass bushing, no ball bearings anywhere inside.
Resolution: Movement every two trim clicks (same as Futaba S3003)
Weight = 58 grams (without fixings or horn)

OVERALL:
I think these should be regarded as high torque, metal geared STANDARD servos. But with poor free-play and no ball bearings. So for applications where high torque is constantly required, the bearings in these are not good enough. The free play can be expected to increase slightly with use and the plastic surround wears.

Construction:
Brass gears that seem well made. No glue on soldered connections, vibration resistance an issue? Motor has worked well. I tried running these for 20 minutes under a high loading, the casing around the motor reached 60°C but steadied at this temperature. The servo still worked well after this.

Fixings:
Servo horns are thicker than usual, and a normal clevis won't clip shut because of this. The shaft will take Futaba horns.
The brass eyelets are TOO LONG! They stick out by about 1mm from the rubber grommets and hence will cause too much movement in the servo on the mounts.
The mounting screws are too short! They only stick out by 4 mm from the brass eyelets!! Nothing like long enough to securely hold the servo, especially if you are intending to use the high torque these servos are capable of.

I'm using one for the rudder with pull-pull wires (nice and tight so the shaft slop doesn't matter.

Would use these instead of standard Futaba servos, if it wasn't for the wobbly servo horn shaft.

John Wh27/10/2016 00:39:34
avatar
12 forum posts
7 photos

Torque test:
I built a test rig with a heavy weight hanging from the end of a 4cm servo horn. As the servo turns the torque builds up and the mass and angle at which it can't turn any more tells us the maximum torque. After several tests with 3 of these servos the maximum torque is 10kg.cm @ 6.6V (high discharge battery - no voltage drop so this is the servo limit and not the battery.

Speed Test:
This is very hard to accurately measure, so I just tested it against a Futaba S3003 servo. These MG996R servos are slightly slower in a 'race' test between maximum deflections.

Freeplay:
Rotation free play is approx. 0.7°
Vertical play on servo horn is 1.4° - this is VERY BAD for a servo. Would not use this for an elevator.

Resolution:
The resolution is not good enough to move for each click of the trim on my transmitter. In general, it only moved every two clicks, but it did move by two click amounts, so it the electronics are registering the new position angle but mechanics can't respond so precisely. BUT, this is as good as the new Futaba S3003 I compared it to.

John Wh27/10/2016 00:52:53
avatar
12 forum posts
7 photos

dsc01894.jpg

dsc01899.jpg

Torque testing ! :

dsc01882e.jpg

ken anderson.27/10/2016 09:17:11
avatar
8791 forum posts
814 photos

I paid a tenner for these and for the price they seem to do the job ok...

ken Anderson.......ne...1... 996R dept.

Tomtom3927/10/2016 10:47:51
697 forum posts
1 photos

Thank you for the info on the MG996R . Very informative.

Funnily enough , I watched a club member thrash his new pride and joy of an ARTF yesterday afternoon , having spent £280 on the airframe plus a new Saito 120 (ish) and five "budget" servos . Post crash study pointed to a failed servo . I've always wondered why people are happy to spend hundreds on a model and yet skimp on the one item that helps it to be controlled!

Bob Cotsford27/10/2016 11:00:15
avatar
8845 forum posts
496 photos

I've got to say that I've used these servos (and the HXT badged version) for several years and found them to work just fine, as good as my Savox units at the very least. I've not noticed any wear develop and I've never had one fail in about 7 years of use, which is more than I can say for Bluebird high torque servos. They may possibly be less efficient than some brands, using more current to achieve the same torque, but not enough to cause any problems so far.

Sorry Taj but I have no qualms with using these particular budget servos.

Tomtom3927/10/2016 11:11:32
697 forum posts
1 photos

Hello Bob,

Sorry I should have mentioned the "budget" ones were not the same as these. Just found out that they were the Bluebird ones from HK. I'm sticking to my old linear Digifleet ones!devil

Geoff S27/10/2016 11:23:56
3902 forum posts
57 photos

Isn't this exactly the sort of thing the magazines should be doing? Critical independent testing of the relatively inexpensive components we're fitting in our models. Servos, receivers, plug in modules, budget brushless motors are cheap enough to buy and test rather than be supplied free for review. And, of course, if they're good enough, the reviewer still has them to use in a model.

Interesting review. It's surprising to me that advertisers can claim something that is so easily verified like ball bearings. In fact it's illegal to do so.

Geoff

john featherstone 127/10/2016 11:33:14
44 forum posts

I use those servos quite a lot I think they may be made by different companys if you notice some are TOWER PRO some are TOWARD PRO note the different spelling I have a lot of these servos all have two ball races in them I did buy 4 from some were in Birmingham that were advertised as having two ball races having opening them they were found to have plastic bushes just like you say so I sent them back I also notice that the servo you've pulled to pieces says PROSTER on it there are that many clones out there so your servo could be a clone of a real tower pro servo or a clone of a clone???????

IanN27/10/2016 12:00:50
avatar
1675 forum posts
119 photos
Posted by john featherstone 1 on 27/10/2016 11:33:14:

there are that many clones out there so ....... could be a clone of a real tower pro servo or a clone of a clone???????

Absolutely.

Nb I'm NOT saying the tester's are a clone. However, if anyone does an internet search and buys servos branded as MG996R direct from China for £4 each - which you can do with these - don't be surprised if they're not the real deal. Known and trusted suppliers are your friend

Jonathan W27/10/2016 13:10:51
133 forum posts
11 photos

I have a "Towardpro" MG996R purchased from Hobbyking and it does have a ball bearing. It also has the solder joints for the motor wires and external lead reinforced by glue. Just goes to show there are variations on what is superficially the same thing.

Martyn K27/10/2016 13:24:18
avatar
5119 forum posts
3714 photos

Thanks - this is really interesting stuff

<snip>

Freeplay:
Rotation free play is approx. 0.7°
Vertical play on servo horn is 1.4° - this is VERY BAD for a servo. Would not use this for an elevator.

What do you mean by vertical play - is this vertical in a linear direction or rotation play centred on the servo spindle? (I appreciate that the implication as its angular means its rocking slightly but would like that confirmed please)

Martyn

Bob Cotsford27/10/2016 13:30:24
avatar
8845 forum posts
496 photos
Posted by john featherstone 1 on 27/10/2016 11:33:14:

I use those servos quite a lot I think they may be made by different companys if you notice some are TOWER PRO some are TOWARD PRO note the different spelling I have a lot of these servos all have two ball races in them I did buy 4 from some were in Birmingham that were advertised as having two ball races having opening them they were found to have plastic bushes just like you say so I sent them back I also notice that the servo you've pulled to pieces says PROSTER on it there are that many clones out there so your servo could be a clone of a real tower pro servo or a clone of a clone???????

Well spotted John, I missed that completely! I have never seen one not branded Toward Pro so I'd say these are definitely clones. I thought mine had ball races and glue braced wiring but didn't like to say without stripping one.

Definitely one to watch for!

IanN27/10/2016 14:56:16
avatar
1675 forum posts
119 photos

I asked TowerPro - their response appears to confirm that TowardPro are genuine. Response is

Hi,
Yes, we do have another brand name "TowardPro" for different market.
Regards,
 
2016-10-27 19:23 GMT+08:00 :
Hi,

I've noticed some servos sold by some suppliers that are said to be Towerpro but on closer inspection the items advertised are badged "TowardPro"

Here's an example

default/towerpro-mg996r-metal-geared-standard-si.html >**LINK**

Can you confirm if servos badged as "TowardPro" are genuine TowerPro items, and if so what - if any - is the difference between those and the equivalent TowerPro badged items

Regards
Ian
 
Capt Kremen27/10/2016 15:17:10
avatar
410 forum posts
169 photos

Agree with GLs earlier post, thorough independant testing should be done of servos, radio gear, models etc. by our model magazines.

Doubt they will though, too much advertising revenue at stake, most, (not all!), reviews are little more than the manufacturers PR copy recycled. And the publishers wonder why there are so many unsold copies left on the shelf at the end of the month.

Bob Cotsford27/10/2016 15:17:59
avatar
8845 forum posts
496 photos

I seem to recall reading that it's because of brand name issues and confusion with Tower Hobbies products in the Western world.

"Proster" on the other hand is a completely new one to me.

Whisperit - can you tell us where those units came from please?

Edited By Bob Cotsford on 27/10/2016 15:19:36

John Wh27/10/2016 17:48:43
avatar
12 forum posts
7 photos

The servos I tested came from an Amazon seller called Proster, or Proser or Prozer (it's actually very easy to change your selling name on Amazon, I think they've been 'developing' theirs!). A lot of the similar looking things sold on Amazon are actually the same and from the same factory (in China somewhere), but distributors buy their own Barcode numbers and print theitr name somewhere on it and try to pretend its unique.

So, from your comments and a bit more research, mine are very likely clones. Perhaps Towerpro didn't trademark the MG996R name, or perhaps the clone distributors are trying to get away with it. Looks like Towerpro's have Aluminium shafts now.

[by "vertical play" I meant the play in the vertical servo horn shaft expressed as the angle of up/down movement this causes in the servo horn. How much movement this causes in your connected push-rod/snake depends on how long your servo horn is.]

Never actually tested a servo before, trying to think of what else to do.

Paul Marsh27/10/2016 19:07:43
avatar
4112 forum posts
1245 photos

I bought a 996R from China. During test it caught fire. Board burnt out. The "Towardpro" ones are ok, they are ok but not very accurate. Although powerful, they drain the battery quickly and someone had two on a Y lead, and the model crashed, due to interference via a Y lead.

Steve Webb Models advises against using a Y lead for this servo (995 Servo)

http://www.servoshop.co.uk/index.php?pid=MG995R

Edited By Paul Marsh on 27/10/2016 19:08:44

John Wh27/10/2016 19:20:26
avatar
12 forum posts
7 photos

Thanks Paul. The servos I tested were each drawing 2Amps when taken to their maximum of 10kg.cm torque loading. The temperature never got above 60°C on the outside of the case (measured using a IR temperature meter).

I always test servos now before putting them in my aircraft. You can get cheap servo testers which will constantly cycle a servo through its full movement . Every servo I get is run for 20mins or so with some loading on it to check it's not going to burn out.

If we're prepared to buy cheapo equipment nowadays (I am!) then I think perhaps we have to take responsibility for testing everything ourselves. Futaba would be concerned if one of their's burned out, can't imagine the clone companies caring so much...

Edited By Whisperit on 27/10/2016 19:21:57

Bob Cotsford27/10/2016 19:27:20
avatar
8845 forum posts
496 photos

Paul, the 995 was a very different beast, badly designed with a very poor amp.

Whisperit, one good test would be putting the servo under load and recording the speed and current compared to something like a Hitec HS645, both for static and dynamic loading with perhaps a further test after something like an hour on a servo test cycle (with a load attached) to show any wear or change in performance.

All Topics | Latest Posts

Please login to post a reply.

Latest Forum Posts
Support Our Partners
CML
electricwingman 2017
Subscribe now
Sussex Model Centre
Slec
Advertise With Us
Sarik
Latest "For Sale" Ads
NEW POLL - has the pandemic altered your event safety perceptions?
Q: Has the covid pandemic deterred you from attending shows and events in 2021?

 No, I'll be attending just as many as I usually do
 No, but I'll choose my event with greater care
 Yes, I'll attend fewer events going forward
 Yes, I wont attend any where previously I have

Latest Reviews
Digital Back Issues

RCM&E Digital Back Issues

Contact us

Contact us

Magazine Locator

Want the latest issue of RCM&E? Use our magazine locator link to find your nearest stockist!

Find RCM&E!