Towerpro Servo Jitters

[Flaps 30]

For a fair amount of time I have found the Towerpro MG90S servos to be pretty good, along with the slightly bigger MG90. It seems that a lot of duff MG90S or poor copies of them have flooded the market. One club member fitted four of them and they all suffered the jitters when you flicked the joystick and sometimes it would happen on startup of the RX. Not good!

So it came as no surprise when I purchased four MG90S servos from a UK supplier (not Ebay) to find that all four suffered the same problem. regardless of what the power source was or any voltages between the normal 4.8 Volts and 6 Volt range ... Quick check with an older MG90S proved that all was well with the system. Below is the end result of allowing the jitters to continue or if you are not aware (as in the fact that the model might be in the air) that the servo is having a jittery fit... That black line on the Y2 transistor on the right is the area that cooked up, letting the magic smoke out that kept it all working.

So it looks like Towerpro servos are now off my list as trusted servos. Now looking at Corona DS939MG servos as a possible replacement. Anyone used them?

[Rob Thomson]

I use the 939 and 929 metal gear versions.

Never had an issue

Sent from my GT-I9300 using Tapatalk 2

[kaos]

used Corona DS939 and DS929 for 3 yrs. Both have more torque than mg90S or mg90 (I only used mg90 which has more torque but slower than 90S). work very well. especially 939 which is digital and has more torque than 929. torque on 929 is on par with mg90 but slower. I think for the price difference I would just get 939. You will find they are a lot quieter than towerpro too. use 6V BEC with them you will like it.
as for the jitter from towerpro, I am not sure it is all about the circuitry. I have used quite a few towerpro mg90 for my heli cyclic before start using 939/929 (never used 90S, torque is weak while not quite fast enough for tail servo) I think part of that is signal noises interference. Some times I switch a different ESC or add a ferrite ring will get rid of it.
I still have some mg90 servos on one or two of my heli, they are fine. I do have about a dozen case/flange damaged towerpro mg90 (by heli crashes ) in my box, now I use them for planes. their metal gear is very strong/resistant to crashes. For all the mg90 servos I have, most of the time it is the case got damaged, not the gear/motor. I think the case will break before the gear/motor.
If high torque is not much a deal but speed is , check out the MD922 and DS919 both has similar torque as mg90S. Both are quieter than towerpro too. MD922 is the smoothest one. among all these. I can use MD922 for either my 450 cyclic or as tail servo.

PS: My Skyfun delta wing has two DS939 in there, each balsa aileron is one foot long 1 inch wide. They can handle anything I put on the aileron.

[Flaps 30]

Thanks guys.. All good information there Kaos.. Ordered a few DS939's and the smaller Corona analogue servo. Let's see how they go. I will push them a bit along with testing them from 4.8 to 6 Volts. Will report back on them.

I do miss the good old days where we had bridge transistors in TO92 packages or E line, where it was usual to have devices that could stand 2 Amps of continuous current and able to dissipate one Watt. Before that time we had centre tapped batteries that simplified the output stage. ...

[MikeB]

Yes, with Germanium AC127s and AC128s. Couldn't let them get hot!

Mike.

[jhsa]

I still remember those..

[Clivew]

Galloping Ghost with Mighty Midget motors, relay driven, then 10 channel reed banks

(I can remember some reed flyers smoother than some of today's pilots)

Manual PPM it was then..had to fly with thumbs.

Those were fun days!

[ShowMaster]

What I hated about the center tap battery pack is of you flew a left hand or right pattern too long you're run out of trim as one side of the pack ran down. You did know when it was time to recharge however.

[Flaps 30]

True Showmaster. Bit like the tyre wear issue on cars that drive around racing circuits in one direction. DEAC button cells were the order of the day. IIRC they had a capacity of around 250/300 mAH

Ahh.. Yes I remember the AC128. In fact I have a few in my vintage transistor and diode box. That transistor was used in a Vidicon camera as the field output drive transistor to the coils around the tube. Never had a failure on the production line from them.

Cooo... GG.. Now that is going back a bit. I always wanted to buy the RCS six channel proportional radio. Sadly it was and remained outside my price range.

As for servo's of the era. They were big things back then. I do re call some red ones that had a screwdriver slot, that you could use to change the servo centre.. Nice idea was the servos that had variable capacitors instead of nasty pots as position sensors. Now who made them? Was it Kraft?

[Clivew]

Heathkit did those I think?
Or was it them that had permeability feedback, variable inductor?
I thought that was better, no contacts at all!

"Now who made them? Was it Kraft?"

Sprengbrook??

[kaos]

Dang! I don't know I am mingled with a bunch of old guys/ladies around here.

Flaps30: Forgot to mention one thing. I don't know how your servos are mounted on your plane. If you use only the flanges to fix the servo, DS939/929/919 case flanges are funny. underside is hollow and it comes with grommet and a metal tube to mount it. those grommets never work well for me (they move under stress). I fill the underside of flanges with epoxy then use washers (or something like this http://www.hobbyking.com/hobbyking/stor ... duct=14941) on top of flanges to fix the servos. Without the rubber grommet, just using screws you can split the flanges by just over tightening (because it is hollow underneath). But if your servo is installed in a cavity then some hot glue or just a dab of contact glue on one side of the case will do it.

[Clivew]

kaos...be very very careful who you are calling old..I am and admit it

[Flaps 30]

Dang! I don't know I am mingled with a bunch of old guys/ladies around here.

Watch/read and learn little sprog.

I do use the hot glue method to install the servo's. I don't fly in low temperatures that tend to make hot glue brittle, so no problem there.

Permeability feedback.. Hmm I do recall some manufacturer trying that in the distant past. I don't recall Heathkit doing the variable capacitor system, although they did use the four wire servo system.

[ShowMaster]

Read and learn Grasshoppers!
The capacitor servos were an attempt to eliminate feedback Cermet pot element damage from engine vibration. About that time plastic element pots came along and I think the caps were phased out.
In that same era another system using a ferrite slug in a tube wrapped with wire forming a inductive tuned circuit. The ferrite slug was attached to the output arm to give feedback position.Many servos had both the rotary wheel we use today as well as 2 linear output tabs.
Oh, in those days all RC systems were made in the USA, not Japan, later Malaysia, and now China.
I restore old RC systems for my other RC hobby and have quite a collection of schematics and manuals on vacuum tune to present times. Even a 5 ch single stick Heathkit with cap servos. I plan to put it on 2.4g at some point.

How many remember one push for right, 2 pushes for left and 3 for up elevator?
Trick question! Answer it and risk being called an old Fa_rt!

[thomas9x]

Phil Kraft (or one of his engineers) did the variable cap servo design. The variable cap looked the same as what you would see in the tuner of a portable transistor radio of the day. I have an old Heathkit R/C system that uses these servos and some still work (but nasty centering/drift).

F&M offered a variable inductor based servo.

Both technologies are probably better at being marketing hype than practical solutions to save us from those evil pots. Just because it's different doesn't mean it's better. Maybe the hall effect designs that have recently appeared will be more impressive solutions.

How many remember one push for right, 2 pushes for left and 3 for up elevator?

{Hand sheepishly goes up.} My first setup was a Kraft single channel tone transmitter with matching superhet relay receiver and two Royal servos. One servo for rudder (right, left, center) and the other for 3-position throttle. Had to click the big button like a madman to change speed. Then I built an analog proportional single channel system (DiY) and that was living the dream.

- Thomas

[ShowMaster]

F&M, yes that's the brand I was missing.
Funny thing is that in those days I think there were less controlled crashes. Many planes were pretty much built to free fly if no control was given. Many flew away because of the poor radio links but when you could find them they may be sitting on there wheels intact, that or in a tree.
Now it's more lawn dart if control is lost.
I guess that's progress.
Phill Kraft would fly at our field as did Joe Bridi of Kaos plane fame. John Perry of Perry Carbs also tested his carb designs and sold us prototypes at the field.
Jerry Pullen of Kraft and PCS design fame lived 20 miles from our field so he would fix our Krafts while we waited.
Jim Odino of S&O custom radio fame still flies at our field but no longer builds RC but is still active in pattern contesting. I enjoy discussing his radio designs and the RC past with him. Then their were these two engineer brothers that were fooling around with electric planes. We all watched the 2 min flights of underpowered heavy nicad brushed motor planes they flew. Very nice guys but we were nitro flyers so we chuckled a little at the so so flights. They got the last laugh years later when they started a company called Astroflite. No more chuckles then, we all bought their brushless inrunner motors and esc's and got into electric flight.
Having these guys around you as mentors gave many of us a much better RC education.
Of course with the forums like this one we all tap a lot of years of knowledge now. The eye ball and hands on sharing of knowledge is what's missing.
To do that we'll need to start Skypeing or FaceTime or some other round table video conferencing so we can share our skills in real time. A lot of scary looking people I'll bet? I see one every morning in the mirror.

[thomas9x]

That's awesome that the prestigious R/C pioneers were regulars at your field. I didn't have access to any R/C mentors so it was all trial by fire.

In regards to the original discussion, do the dicey MG90S servos sold today have a different electronic design than the old "good" ones?

- Thomas

[ShowMaster]

From what I have seen yes, but that expected. Pretty much one IC doing all the work. Some have FET output motor drivers that can really pump the current into the motor.
Them there's the newer digital programmable servos.
The biggest complaint seems to be a bad buzz at test. Usually this is because the dead band is very tight and the servo constantly tried to center itself and can be normal for that servo, If the pulses from the receiver are a little less square, there edges tapered then the servo doesn't know where to stop.
That seems to be on the less expensive servos as well as a feedback pot quality.
The older servos had a lot of deadband so they never buzzed and also didn't trak position or center perfectly all the time.
Usually the sticks were as bad.
I see so many post about jittery or buzzing servos these days. Usually in $5 servos.
I'm guilty of using them and for the price I buy extras and select the best of the lot. The others are used in a pin e or for spare parts.
If you find a brand and price that worked well with you radio system stay with them and post their model number so we can try them.
I use a lot of HS 55 servos and they do start to skip or stuffer after many flights. I've had a few just burn up after bring powered doing nothing.
So now I made a quad servo plug cable for my servo tester and auto exorcise my new servos for 30 min with some weight on the arms. Then I install them.
It sure sorts out the new bad ones. Better on the bench than in flight.

[G550Ted]

"auto exorcise my new servos"
I just dunk mine in holy water and let 'em drip dry. If they still work after that then I know they are good to go. The ones that don't work were obviously posessed so I burn them on a stake!

Ted

[Flaps 30]

The servo controller chip that is in the buzzy MG90S servo's is the KC2462. Unable to check what the good ones have, as I have none around that are not installed in an aircraft.

Thanks for the servo information Thomas. I thought it was Kraft that did the capacitor servo. Nice to see the cost of the two servos you posted, $40 a pop way back then was a lot of money. We are not paying enough for our kit these days.

[kaos]

Ordered a few DS939's and the smaller Corona analogue servo. Let's see how they go. I will push them a bit along with testing them from 4.8 to 6 Volts. Will report back on them.

I have been using DS with 6V on the heli from day one, never had a problem. on the skyfun the BEC is 5.5V, no problem either. when you use them as tail servo, they do get warm/hot cause it is working constantly. but on a plane, it is barely warm at all.

Watch/read and learn little sprog.

yes, Ma'am!

kaos...be very very careful who you are calling old..I am and admit it

yes, SIR!

[Clivew]

That's better, show respect for your elders

"Oh, in those days all RC systems were made in the USA, not Japan, later Malaysia, and now China."

Not quite right
The Flight Link analogue proportional systems were designed and manufactured in the U.K.
Started as 2 channel superregen 27mhz, went on to become 4 channel superhet on 35mhz.
Had a brazed aluminium transmitter case. Wasn't as susceptible to interference either, as
it transmitted continuous tones.
There were plenty of other european manufacturers too, Graupner, Simprop for a start

[ShowMaster]

Then of course there was the cohearer radio control system. No tubes, no transistors, and yet it worked!
Off servo topic but looking it up will make us all appreciate what we have today.
I actually repaired a cohearer RC bus toy and got to play with it. Amazing technology for its era.

[Clivew]

Wow!

How did you get the spark transmitter certified

[ShowMaster]

Talk about broad band!
No gold sticker for that transmitter

[Flaps 30]

Then of course there was the cohearer radio control system. No tubes, no transistors, and yet it worked!

This is what early radio control was about. Here is an interesting site that you wrinkly folk might like to explore ---> http://www.mccrash-racing.co.uk/sc/early.htm

My first introduction to RC was in seeing a German RC system, where the RX was one module with another module stacked on top. I do recall seeing that the RF IF section had a Valvo OC45 in place. I wish I could remember what the name of this system was, it was top of the line stuff as most RX's were super regen's and this system had a superhet RX..

This is the pub, so we are allowed to drift around a bit, and it amazes to see how many ancient folk who are here living out a dream, before they end up in some rest home.

[Flaps 30]

Ohh.. You do know that you can make a good radio detector out of a piece of galvanised steel... Not so good for RC but real fun to get working.

[jhsa]

I have this one

http://wiki.rc-network.de/index.php/Bec ... 9Cbersicht

João

[Clivew]

Ohh.. You do know that you can make a good radio detector out of a piece of galvanised steel... Not so good for RC but real fun to get working.

Yep!

Now how do I put this without being handbagged

I didn't know there was anyone as old as me on here

No, I'll get handbagged for sure

[MikeB]

I do re call some red ones that had a screwdriver slot, that you could use to change the servo centre

Sprengbrook servos (some red, some grey) had a screwdriver slot to adjust the centre. I used them for years, nice linear outputs, masses of force (not torque as they are linear!). I've still got some, but the cermet pots are worn out. Like many servos of that era they have the 5-pole motors.
They also mounted in metal clips, in a hard arrival they would spring out, rather than become damaged.

Mike.