Assan GA250 MEMS Gyro

[exblade]

For those interested in a really good $10 MEMS gyro, there is a thread going (up to page 84 since January) over on RC Groups. I have been fooling with them for a while now and highly recommend them for a nice inexpensive gyro. I have been using the 401B style gyros for a couple of years now and prefer the GA250. It has a MEMS sensor compliments of the mass production of chips for iPhones and iPads, and the avcs software is similar to the Quark. The gyro holds the tail like glue, so for hovering and learning and sport flying it is unreal. You need a decent digital tail servo on it like a Henge 922MD or DS480 as the servo must be at least 250 hz frame rate. Below is my latest post on that thread, having flown the King 2's yesterday after work.

Well guys,

I finally got to fly my two identical modded up King 2's both with Typhoon 2215H motors, 11 tooth, 450 V2 heads and Henge MD922 tail servos. One has a GA250, one has a HK401B. I ran two batteries through each one, alternating. The wind was only moderately gusty, say 10 - 12 knots. After flying them back to back, here is what I think.

1. Either gyro with this tail servo is MORE than good enough enough for beginners through sport flyers. Edge to the GA250 for zero drift.
2. The HK401B is easier to setup and the AVCS software seems to work slightly better. Not a hint of tail bounce for example.
3. The GA250 holds the tail better, staying right where you left it.
4. I prefer the feel of the GA250 better. Better piro rate, even at 75 endpoints to reduce bounce to a minimum. The feel is more natural in banked turns and the like. Not a huge difference, but enough to do the upgrade. And if your are like me, it will force you to get a better tail servo which makes a BIG difference. The HK401B with the 922MD is much better than it was using an analog servo.

I am still running one of the late January (2nd batch I think) GA250's and need to change it to one of my more recent ones, but the GA250 is a keeper. BTW, run the Henge at 250 hz and it stays happy.


Here is a link to the thread:

http://www.rcgroups.com/forums/showt...362819&page=84

The thread is pretty interesting as Assan responds to many of the postings and has been taking in the suggestions. There is a setup video out there that someone made that is quite excellent. I found that on the King 2 with the 922MD, I needed 75% rudder travel to keep the tail from bouncing and about the same gain as I was running 76 in HH on the DX6i. The tail slider is centered at neutral.

Here is the gyro:



http://www.hobbyking.com/hobbyking/s...dproduct=15318


Here is the setup video by ghost2212:

http://www.youtube.com/v/KRbpVmcvo7M?fs=1

Enjoy.

Terry

[likeakite]

Got two on the way. Enough good reports and the inexpensive price make it hard to not try out. I just put a Dynam GY48V on the stock k-3 and havent had the chance to put it thru its paces yet, or do any tweaking, but proly at some point, will yank the GY48V and install the GA250, along with one of the tail servos that seems to work well with it, and see what GA250 can do.

[exblade]

Hey Kite,

I have two of the GY48V's flying right now. They work pretty well, especially with the DS switch set to ON and a digital servo. I will be interested in what you think since I think you have a GY520(?).

I just got a Quark and MKS 8910 (760 u 560hz) servo for my 500 to replace the GY48V. (I guess I went crazy one night.) I am going to fly it as is for a while and then switch to the Quark. It has a Blue Arrow 1520u 333 hz tail servo on it now and works pretty nice with the GY48V. That servo will end up on a 450 with a GA250. One thing to mention, most of these tail servos REALLY do not like going over 4.8 or 5 volts and burn up. A very easy fix is to just solder a 1 amp diode in line with the red connector. The diode (radio shack 99 cents) drops the voltage .7 volts with no other ill effects or complications. Nice protection for your servo, and you can run slightly higher voltage for your other servos.

Give the GY48V some flights to see how you like it. I have a sneaking suspicion that dynam has been making all of these 401B's and the 48V is their own branded one. They are definately more sensitive to vibes than some other gyros, so I always use two pieces of 3M outdoor mounting tape.

Enjoy,

Terry

[grawyl]

Hey, exblade.

I just put one of these on 450Pro today (replaced a GY48V which I hate because it drifts like crazy). I also have another on the way to put on 500 and I am hoping I get this all set before it gets here.

Well, the GA250 holds quite well (no drift) but I have a nasty bounce. You mentioned that you lowered your endpoints to 75. How much diff did doing this make? I will try doing this tomorrow but I ask now because I am curious and its dark outside so I cant test right now.

I've been working on getting this thing setup most of the day and I have reduced the bounce quite a bit but its still beyond my liking and if I turn my gain up above 52, the bounce will continue in a fast wag back in forth (very fast bouncing back and forth about 60 degrees).

I am using a DS3400G at 5V (333hz) with the ball about 8mm out on the servo arm (Started at 10mm), gain is at 51 and end points set to 100 in each direction. Also, I am using a 1mm foam tape, the weight, and then another 1mm foam tape.

[exblade]

grawly,

The end points make a large difference in the bounce. Also, I have mine setup with neutral in the center of the tail shaft, not setup for rate like I used to do with the G401B's. Also, closer to the center on the servo horn helps. ghost2212 has a video showing his Trex250 working pretty well with the GA250 and has his radio setup showing on the front of the video, you might take a look at that. The 450 should be easier than the 250. Here is the link:



I don't know about that servo, other than it being pretty expensive. The Spartan site shows that one with 2 stars versus quite a few others with 4 stars, so there must be some issues with this type of AVCS setup. I just got a MKS 8910 narrow band (760u 560hz) servo to try out and it shows as a 5 star on the Spartan table. BTW, this is a quite handy table as I can never find servo frequencies, especially for JR. Here is the link:

http://www.spartan-rc.com/resources/servodb/servo.php

Another thing you might find is that you need a different endpoint for each direction as the bounce will differ.

As a side note, the bounce is being caused by the deceleration rate and gain (two AVCS parameters) not matching the physics of the actual helicopter. Spartan allows you to access these using their USB software. If we could do this on the GA250, I think it would be perfect.

So the endpoints modify the piro rate which can help with this mismatch. Keep in mind that the endpoints in the transmitter have NOTHING to do with the travel of the servo, just the piro speed. You set the servo endpoints in the gyro during setup. Also, anything you can do to make your tail smoother also helps.

Terry

[grawyl]

Thanks Terry,

Well, I went out and tested it today and I was not very pleased. I moved the ball to 6mm from the center and I adjusted the EPA to 80 in each direction.

I was able to increase the gain but It would not hold the tail in this configuration. On hard pitch the would rotate about 60 degrees around and I still had some bounce.

It would however hold in the wind quite well though. The wind was blowing about 20mph.

I have taken the Jr servo out and I will try a gotech. If that works, I will either order another gotech or an EMAX ES20D from ValueHobby because the specs look pretty much the same.

[grawyl]

I got a short test in this evening. It s blowing 20-30mph but I was able to get a few min flight in to see if the goteck made any diff and...Yes it did!

I moved the ball back out to 8mm on the servo arm and I have the gain turned up to 65 (so far). It was holding very well with a side wind and the tail bounce is only visible upon stops from a left hand piro. I lowered the EPA on the left side and that seemed t help some more.

I have learned that without software or a POT to dial in the delay, this gyro requires a fast servo (.07sec/60 or better).

So my last question. Which requires a faster tail servo; larger heli or smaller heli?

[exblade]

grawyl

No clue on the question, except my answer would be BOTH. It would take a pretty wild simulation model or physical test to decide. If you want some good info, read the Spartan RC knowledge base. They impressed me so much, I ended up buying a Quark. Primarily to have a benchmark for these other Gyros. I have yet to mount it though. I did have a look about with the PC connection and looked at all the AVCS parameters.


Terry

[grawyl]

Thanks Terry,

I went ahead and bought another servo. It should be fast enough and strong enough for the 450 and I will put the gotech back on the 500. Really appreciate the help tho.

[Fundafly]

This might sound confusing but it does work.

To get rid of the bounce try shortening the distance between the servo output ball link and the center of the servo (7 to 8MM), and move the ATV back to 100%

If you just cut back on the ATV this does nothing more than limit you travel, and does not help to increase the resolution of how the gyro detects the tail/servo movement.

By moving the ball link closer to the center it allows the gyro/servo combination to have a finer resolution to work with.

You have to have a servo that will work with the frame rats as spec'd by the gyro mfg.
I have been using the GA-250 with a MKS DS480 servo in a 450 with very good results.
I did try it in the configuration you have tried (just cutting back on the ATV) and I did get a fair amount of tail bounce.

As for a bigger bird I don't know if this applies.

[exblade]

Fundafly,

Thanks for the note. All the setup info talks about putting the servo ball in at the inner hole as is common on a 250. Mine is about 6mm on the King. How far out is your 450?

Beyond that, the ATV or endpoints reduce the piro rate, not the limits (in HH). I think this helps by reducing the rate the gyro tries to make the heli spin and then decelerate. It is this inability of the heli to stop as fast as the gyro wants that causes the bounce back. The heli decelerates the tail but goes past the stop point and then must be corrected back to the gyro heading. People are reporting good results on 450's but the smaller heli's are having problems, although there a few guys with 250s that are happy and a few 450 guys have given up. A few guys reported similar issues on Spartan's which are tuned for 500 and 600 size heli's. I just put a Goteck 9257 on my 450 and will replace the GY48V with a GA250 in a week or so.

I have been flying my King with the GA250 on it and am very happy with the performance. Flew four batteries today in gusty winds and it was flying better than ever. When I get around to it, I am going to put my Quark on an identical King 2 and fly them back to back. It has a HK401B on it now and the GA250 is better. I am actually expecting some bounce on the King with the Quark, though could be wrong. But with the USB connection and software, I can see and adjust what I need to on the Quark. I am hoping Assan will make some changes so the AVCS parameters are accessible via USB, then we could dial away any residual problems. As it is now, with some tweaking it is a great low cost gyro.


Terry

[exblade]

Well guys,

Today I moved my Quark over to one of my two identical modded King2s. Both using the Henge 922MD tail servo running at 250 hz. Ball at 6mm from center. The Quark out of the box is working a bit better than the GA250 on basic driveway bouncing around and doing piros. I am going to put a later model GA250 and move the ball in to 5mm on the other one and then see. I also need to do some open space flying as pumps and piros don't really tell that much. I did have some tail bounce on the Quark, but only occasionally and I have not settled on a final gain yet. As far as hover tail stability, I would say no difference between the two. These gyros I think will both do better with a 760u 560 hz narrow band servo like the MKS 8910 that I bought for my 500. The Quark is going on the 500 with the MKS.

Terry

[Fundafly]

Both of the birds with the GA250 have the ball link 7MM from center, and all is working good.
The only reason I mention about the ball link distance is that when this gyro first came out a lot of the guys were saying that they were having a lot of bouncing going on, and then all they did was to move the ball link closer and this solved the problem.

The only thing that I could see is happening when they do this is;
A. they get a little faster reaction in regards to the servo movement
B; they get better resolution due to the shorter servo arm sweep, and with better resolution the gyro can control the servo with better accuracy

So far I am happy with this gyro, it does a better job than the HK401Bs that I have been using.
It is holding well in all flips, FFF, FBF, and FFF flips with max + and - pitches during the tumbles.
I am starting to learn the more advanced 3D crap piro flips, etc. so we will see how it is there.
I have never tried a Quark, or any of the top of the line gyros yet, and don't really plan on it at this time, as the Assan appears to be working well at this time.

Happy Flying
Tom

[exblade]

OK, a few more batteries in both Kings. I had a small setup issue with the King with the Assan. One ball link on a tail blade was off the ball!!! So I was only getting push, not pull on one blade. Pretty weird. After fixing that and tweaking my gear mesh and REDUCING the gain, the thing flies brilliantly. The Quark is still flying well, but needs some tweaking. BTW, the gain is lower on the Quark than the GA250 68.5 vs 70 on the DX6i.

For the type of flying I do on a small helicopter, the GA250 is all I need. This newer one powers up no issue, is easy to setup and flys great. BTW, my servo ball is 5mm on the Henge 922MD on this setup. I think the Quark on my 500 with the 560 hz servo will be sweet, but the GA250 is going on everything else.

Terry

[polarisflyer]

So when I paid $10 for an ultra small gyro (MEMS), I really didn't have high hopes to be impressed by it. After struggling with my old MKS 292, I just assumed all cheapy gyros held just "OK"....

Today, I finally figured out how to configure the Assan GA-250 MEMS gyro correctly, with the proper digital servo selection, as well as the min / max travel end points, and I was absolutely blow away at how stable the tail is now!!!!! WOW!!!! It's small, it's light, it is faster and ultra steady! If anyone needs even more of a push to give this gyro a try, for $10, you really will be impressed!!!!! TRY IT !!!!!

[likeakite]

Hi PF
Curious about your servo, and what chopper its on? Have two of these Assan 250's but havent had time or need to install one yet. Sounds as if exblade sorta has it figured out for his K-2. Seems as if end points and ATV's play a part in getting a good setup too, from what you guys are experiencing.

[polarisflyer]

Originally Posted by likeakite

Curious about your servo, and what chopper its on?

I have an HK 450 Pro TT with an HK Mi series 28012 Digital Metal Gear

If it wasn't for the fact that I decided to remove the stock landing skids to replace with more robust (read TALLER) 500-size skids, I'd take a photo to show you how nice the installation is. I mounted the MEMS gyro on a sandwich of a metal plate in between two layers of two-sided foam tape, right above the tail boom block at the rear of the frame. This keeps the vibration down to a minimum.

[likeakite]

I have 450 skids on a K-3, 500's on the 450, and 700's on a 600 heli, and think it makes them look better. Sure does help keep the tail blades in better condition and out of the grass and dirt.
That looks like a decent servo, appreciate the link and tucked that bit of info away if needed. The list of things to order from HK might get an addition.

[Jomaba4308]

I am building a TT Mini Titan v1 and could not resist trying the GA250 gyro but I am nit getting much servo travel. I have the gyro paired with a JR DS290G tail servo. Also the light on the gyro is solid blue but when I rotate the heli by hand nothing happens, the servo does move with stick input (barely) but the rotating the heli does not move the servo.
I realize that this thread is in the E-sky section but cannot find a thread for this gyro anywhere else, please help.

[polarisflyer]

Perhaps you need to reinitialize the gyro and change the servo type. I found that I needed to change my servo type to get the best response. Once you have selected the correct servo type, then you need to program the min / max end points. If the servo doesn't move even when you enter the Travel adjustment mode, then you may have some incompatibility. Check the servo itself to make sure it hasn't gotten burnt up with the wrong timing from the gyro, by connecting the servo directly to the receiver and check travel movement...