Improving a Kit Servo Saver

[speedy_w_beans 4272]

Currently I'm in the middle of installing electronics in a few touring cars and pan cars, and one thing that has been really irritating me lately is sloppy servo savers. I have a few of the Tamiya beige/gray servo savers for RM01, F104W, and Group C chassis, and these things are all over the place. Not only do they rock from side to side a few degrees, but they also float a little underneath the horn's washer/screw.

I also have a few black servo savers for TB03 touring chassis, and while these don't float as badly as the pan car servo savers, they do rock from side to side. I've modified the plastic steering rack to use flanged bearings in the past; that helped tighten the steering from knuckle to knuckle. However, the connection between the servo output shaft and one of the steering arms still suffered from servo saver slop. All the ball connectors and ball cups were in good shape.

I guess an easy answer is simply buy a Kimbrough servo saver and call it good. That's at least an extra $4-5 each time I build a kit. Another option would be to skip the servo saver and just run a straight arm included with the servo. That's taking a risk of stripping the servo in a crash.

With this in mind, I took a closer look at one of the TB03 servo savers tonight and improved it pretty dramatically with only some thin strips of Scotch tape. The thought process outlined below could be used with virtually any basic kit servo saver to improve it.

To start I captured a baseline measurement of slop. The servo itself was tight and the saver wasn't loose on the servo output shaft, so I was confident all the measurements would reflect the servo saver itself. With the servo in a fixed position I applied extremely light finger force to the horn to deflect it clockwise or counterclockwise, took angle measurements, and then found the difference between them. I was careful to not deflect the saver spring or change the servo position while taking these measurements.

Clockwise measurement = approximately 20 to 30 minutes to the left of vertical, or about 0.5 degrees.

Counterclockwise measurement = approximately 3 degrees 30 minutes to the left of vertical, or about 3.5 degrees

The total rotational play in the servo saver is about 3 degrees. That's huge!

While rocking the servo horn back and forth, I noticed a few things:

1) The spring captured between the base and horn of the saver floats a little on the horn. The only part of the spring that matters is how it contacts the locating fingers on the base and horn.

2) The locating finger on the base is just a little wider than the locating finger on the horn. This means the horn's finger is free to move between the spring's gap, which is set by the wider finger on the base. The horn then rotates freely a little as a result.

3) The base provides a hollow stub that the horn locates on and rotates on. There's enough of a dimensional difference between the stub on the base and the hole in the horn that additional horn rotation occurs after the locating finger touches either side of the spring.

To get to the bottom of all the tolerance issues, I took some measurements of both the base and the horn.

The base's diameter for locating the spring is 10.18 mm.

The base's diameter for locating the horn is 5.56 mm.

The locating finger on the base is 2.48 mm wide.

The spring gap is 2.06 mm; this is when the spring is relaxed.

The horn's diameter for locating the spring is 9.94 mm. This is 0.24 mm smaller than the base. At first I thought there was something to correct here, but then I realized it all comes down to how tightly the horn is located on the base, and how matched the locating finger widths are.

The horn's diameter for locating it on the base is 5.66 mm. This is 0.10 mm larger than the base's hollow stub. Even if the locating fingers were perfectly matched, there would still be some slop because the spring is slightly oversized and does not perform any locating function between the base and horn -- it all depends on these two diameters being closely matched.

The locating finger on the horn is 2.33 mm wide. This is 0.15 mm narrower than the locating finger on the base. This definitely contributes to the free rotation of the horn as the spring gap is never smaller than 2.48 mm.

So, I was thinking about what sort of material I could use to more closely match the critical dimensions of the base and horn, and I thought Scotch tape might be a good idea. One side has adhesive so the tape will definitely stay in place. The other side is relatively smooth and low friction. Plus, the horn and base never really move relative to each other in normal operating conditions -- the spring with 2.06 mm relaxed gap actually has some preload on it when it is spread to 2.48 mm. This means the servo saver can handle some load before the spring even thinks about spreading a little.

As it turns out the Scotch tape I have on hand measured between 0.05 and 0.06 mm in thickness. This is perfect considering I'm trying to address 0.10 mm and 0.15 mm differences in dimensions.

To address the difference in diameters for locating the horn on the base, I just applied one layer of Scotch tape. This increased the diameter by about 0.10 mm.

Test fitting the horn on the base, it was immediately apparent all the slop from this issue was gone. The horn still rotated freely on the base, but I couldn't detect any play between the two parts like I could before.

The next issue to address was making the finger on the horn the same width as the finger on the base. Since there was a 0.15 mm difference between them, I added a thin strip of Scotch tape to the finger (0.10 mm of width, 0.05 mm on each side), and then added a very small square of tape on one side of the finger (adding another 0.05 mm of width). Total change was 0.15 mm.

Here's the assembled servo saver now. The horn is located on the base with nearly 0 difference in stub/hole diameter, and the two fingers between the ends of the spring have almost 0 difference between them too. The assembly felt really precise.

So what are the results?

Clockwise measurement = 3 degrees 40 minutes to the left of vertical, or about 3.66 degrees.

Counterclockwise measurement = 4 degrees to the left of vertical, or about 4.00 degrees.

Total rotational play in the servo saver is about 0.33 degrees, about a 9x improvement over the original measurement!!!

This was for a TB03 kit servo saver. For the RM01/F104W/Group C beige servo savers, I think the same process can be applied, but I would also look at filing the base's stub a little so the washer and screw hold the horn more closely against the base.

My job is done here...

[Nobbi1977 1793]

I love it when a rational person with skill get pushed to the projects like this :-)

Great work and measurement. I just trimmed a bit of the top of the posts so it tightens down more, now I will use your tutorial.

I have read some people use a blob of superglue. Takes out the slop but will crack under pressure.

[oz-trash-cowboy 63]

Even better, delete the servo saver altogether, fit a digital metal geared servo and aluminiun servo horn and use all your new found steering precision to totally avoid crashing!!!!!

[Ronnyhotdog 221]

On 1/11/2015 at 6:01 AM, speedy_w_beans said:

 

Can you tell me what this measuring device is, please, or where I can get one?

[speedy_w_beans 4272]

It's called a protractor, it can measure both outside and inside angles.

I bought mine at a local (USA) Harbor Freight store, but you can get one on Amazon too:  https://www.amazon.com/General-Tools-29-Plastic-Protractor/dp/B00004T7P5/ref=sr_1_35?ie=UTF8&qid=1520018543&sr=8-35&keywords=protractor

There might be other, less expensive sources for it; you'll have to shop around and look.

I like the fact it can measure inside and outside corner angles, and it also has minute-of-angle measurements via the vernier scale.  You can also lock an angle with the brass thumb screw.  The brass flange tube edges in each joint are rolled and add some friction to the links, so you have to be mindful of that and make sure the edges of the protractor are truly touching what you're measuring.  You might see a third of a degree of error otherwise.

 

[TurnipJF 9192]

I van vouch for the tape method. I used it on my DT-03T build to excellent effect. 

[ChrisRx718 4470]

That's crazy - I guess the tolerances are such to encourage upgrading.

One thing I've always kept in my kit box is a 3Racing servo saver. Virtually the same design as the Tamiya one but only $2.50 USD. im down to my last one, must be time to buy some more!

http://www.rcmart.com/3racing-servo-saver-tamiya-f104f103-3racshamu-p-31562.html?cPath=1099_808

 

[nowinaminute 2815]

On 3/10/2018 at 7:32 PM, ChrisRx718 said:

That's crazy - I guess the tolerances are such to encourage upgrading.

One thing I've always kept in my kit box is a 3Racing servo saver. Virtually the same design as the Tamiya one but only $2.50 USD. im down to my last one, must be time to buy some more!

http://www.rcmart.com/3racing-servo-saver-tamiya-f104f103-3racshamu-p-31562.html?cPath=1099_808

 

Awesome, I've been looking for a half decent cheap one for ages. I do a lot of toy grade conversions and other obscure relics and oddities and it just doesn't make sense putting a £15+ servo saver onto a £2.30 servo in an RC that cost £5 lol.

I've just been buying the Tamiya servo parts split from kits for around £3-4 and putting super glue on them to tighten them up. I'll have to get some of these in!