JR-9303 2.4 Field Results

[GooseF22]

I would zero in on the voltage reset first. There are alot of questions to ask, but the main one is power delivery. Ask him about his setup, look it over, see if any servos are binding, how many plugs on the batt...etc... nothing new.... Its just that this baby is more sensitive to downspikes. When I set up my suky, I used a futuba ppm just to get things going as waited for the first dx7, and took some voltage readings while bangin the sticks, and I noticed a larger swing in voltage with the 7000. I am assuming that is due to the speed, it can bring the servo to its fullest speed because it can reverse the signal so quickly. So my thinking is that the same system could have a larger downspike. So, the regulators become more critical if you run them, ones that have high capacity, and dont heat. Heat is a killer. So do powerful batteries, and good wiring and double switches. It has to do with the load slope on the motors in the servo. Dick said awhile back plan 1 amp per 100oz for good margin - good for me.

Capacity. most manufactures of electronic components advertise peak power. Use 50% of that for max sustained, and 70% for average power with minor surges, but then look at the plugs. no more than 4 amps per connector. For examply my suky with dual miracle switches is supposed be 10 amps continuous, 20 peak.. thats crap. its about 7-8 maybe 9 amps sustained. I figure anything above that could cause a spike out on the AR receiver series. The problem is that the AR even sees a momentary spike down, you could be in trouble. What we need is a stabilizer circuit with a built in capacitor that can smooth out the millisecond spikes like a shock absorber. any brainiacs?

keep the info comin guys.
g

	Quote: Originally Posted by 1bwana1
	Yesterday, we tested the system in a 40% Carden Extra. Once again, the transfer of the programming between the 72 MHz 9303 and the new 2.4 X9303 went perfectly. This time the RX was mounted to a Smart-Fly power Expander. I read somewhere that the Smart-Fly power expander line is the only system that has never shown a reboot during testing at Horizon, so this seems the safest way to go. We got a perfect range test at double the recommended distance. When we range test, we always rotate the plane so that we test from all angles. Three of us flew the plane on multiple flights with perfect results. We all agreed that the plane felt "touchy" so we ended up increasing the Expo in the radio and it settled in perfectly. Could this be the result of the higher resolution? The current setup still only has the main RX and one satellite. The extra satellite RX and data logger have arrived so they will be installed prior to the next test. The connection seems solid but without data, we could be on the edge and not know it. It will be interesting to see the results. There is another guy flying a DX-7 based system (AR9000) at the field. It is in a 2.6m CompARF Extra. He has the main RX and 2 satellite RX installed, as well as the data logger. He has flown many flights with the system. However, on his last flight, the plane went into lockout for about three seconds. Fortunately, it came back before he crashed, and he was able to land without damage. When the data logger was examined, it showed no dropped frames, fades or any other problem. We assume that this means that the RX re-booted, wiping out the data. I always thought that it took longer than that for a re-boot, but with no other indication of problems, what else could have happened? I guess this means that the data logger is not useful for diagnosing voltage issues. This is a problem, as it seems voltage drops are the most frequent cause of crashes with this system. He is now afraid to fly the plane. He did not have a power expander installed, so I have suggested that he get one. Given what we have learned so far, I can see the need for a product designed specifically for the Horizon/JR 2.4 systems. It would include the current power expander features, plus a data logger that includes voltage info that persists after a re-boot. It should also include some system that absolutely keeps the voltage above the re-boot limit (assuming the batteries don’t fail). We will continue to experiment with the system as we escalate it up to the competition planes. We’ll keep this thread updated as we do. In the mean time we are interested in hearing from the rest of you.

[1bwana1]

Here is the report from today's testing, I am afraid this will get a little long, so please bare with me. We are still flying the 40% Carden Extra, because it is all wood, and we are reluctant to install in the ZN-Line Composite plane as this is a more difficult environment for 2.4. Today we had the additional remote RX bringing the total to 4 receivers. So the current setup is:

40% Carden Extra (all wood) This plane has years of flights on 72 MHZ with no issues
DA-150
30 X 10 Carbon Fiber prop
16 JR 8611 Servos with Match boxes
1 Fromeco 2400 for ignition
2 Fromeco 4800 for RX
3 HD switches
1 Smart-Fly Reg on ignition
2 Smart-Fly Regs on RX
All batteries, Regs, and switches have 16 gauge wires and Deans connectors for good power supply
Smart-Fly Power Expander Sport
R921 2.4 RX with 2 remote receivers
Spektrum Data Logger

The main RX is installed on a raised platform that was built to get the RX up high in the fuse and away from the fuel, tank, motor, wires and other things that might interfere with it. It sits in the normal spot on the power expander, which is mounted on the left side of the plane. The first satellite RX is mounted on the opposite side of the plane, and orientated 90 degrees to the main RX. The second satellite RX is mounted about 1/2 way between the canopy and tail. It is up high in the turtle deck, and orientated so that it is different from the other two RXs. We could see no way that all 4 RXs in three locations could be shadowed from the TX simultaneously.

For those who have never used the Spektrum Data Logger, it logs fades at each RX location, dropped frames, and holds. Here is how the data is referenced by the logger.

A = RX-1 in the main RX body
B = RX-2 in the Main RX body
L = Left remote RX
R = Right remote RX
F = Dropped frames
H = holds


After yesterday's good test we were very interested to get actual numbers from the data logger. We did a range test again, which looked good, and took off. A flight of about 14 minutes, including the IMAC Unlimited sequence was flown. Here is the data from the logger:

A = 255
B = 740
L = 509
R = 520
F = 304
H = 1

We were not sure what to expect, but still, we were not impressed with what these numbers were telling us. There are a lot of fades on all RX's, over 300 dropped frames, and 1 hold. It looked to us like we were on the ragged edge of losing the plane, so we decided not to fly any more until we figured out what these number actually indicate about the robustness of the system, and our installation. We realized that we had not re-set the RX between the range test (with the power reducing button pressed), and the flight. It seemed reasonable that there would be fades, dropped frames, and even a hold with the low TX power. So we decided to do more tests.

First we re-set the RX and did our range test again, exactly as before. Here is the data from that:

A = 255
B = 279
L = 275
R = 191
F = 310
H = 5 (much more here this time)

It seemed to us that this could account for almost all of the scary numbers from the first test. We reset the RX and walked the exact range test line again without the range test button pushed so we had full TX power. This time all data values were 0. Very encouraging!

We decided to fly a short 4 minute flight and check the data again. Here are the results:

A= 88
B= 109
L= 96
R= 184
F= 4
H= 0

It still looked like a lot of fades, but because there were so few dropped frames and no holds, we thought it looked like at no time were all RXs blocked at the same time. So we re-set the RX once again, and flew a full 14 minute flight, which included the IMAC Unlimited Sequence once again. Here are the results:

A= 255
B= 796
L= 829
R= 894
F= 132
H= 0

Unexpectedly, the fades were even worse than the first test. There were still a lot of dropped frames, but this time no Holds.

Because we really don't know what kind of numbers indicate a properly working system, we decided in the interest of caution, not to fly with this setup again, until we fully understand what is going on. Now the search begins for information that will allow us to properly interpret this data. We are not giving up on the system, it flew nice. However think about it, without seeing the data from the logger, the range check seemed to go well, and the plane flies great. If this data does indicate that the system is not properly optimized by our installation, we could have very easily lost the plane. There are many out there flying 2.4 without doing this kind of system testing. What are the ramifications of this? Surely, at the very least, people are losing planes unnecessarily. I guess we will know more as our understanding of the system grows.

I would really appreciate comments from anyone with real insight to this data. In the mean time, I am going to forward this report on to JR for their comments. I'll let you know what I learn.

-STEVE-

[1bwana1]

There is one thing that did occur to us. We don't know of anyone who has done this kind of analysis with the standard 72mHZ radio equipment. That might be equally as scary, but who knows?

[walex]

Hello All,
Waiting to get my Seagull Yak in the air, but the weather is starting to change here in Ohio. This is my first time with 2.4. If anyone has any pointers about my setup, please let me know. Thanks.

[Tailwind]

Hi Steve,
I have the Spektrum module in my 40% Carden, set-up is almost the same as yours, I have never seen over 150 on each ant, or 23 on fades, and 0 on holds. After most flights I show 0 fades and @50 on each ant. This is on 15 min flights practicing IMAC, and I have fown it out pretty far. I would move the receiver and antenae's around and see if it gets better...My receiver is on the left side of the fues behind the aft wing bolt. One ant is on the lower cross members in the center of fues, and one is up on motorbox right side by fuel tank.

Dom

[1bwana1]

Walex,
Judging by your pictures, the positions of your main RX and satellite is similar to my friends Carden. If I were you, I'd get the data logger to test as we are. From the feedback we are getting our installation is not working properly.

[1bwana1]

Tailwind,
That is very revealing info. Your numbers are way lower than ours. I'm surprised that you mounted one of the remotes up close to the engine. Intuitively it seem that the further away you stay from large metal objects the better. Did you use the data logger to help position things? I'll try and get my friend to send some pictures of the installation and post them. He and his Carden are serving as the guinea pig for the X9303 tests.

[Tailwind]

Steve,
The ant on the motorbox is on the aft section directly above the wing tube. My digital camera is MIA or I would take some pics of my installation. I have not had to move mine at all, I wanted a good spread, between the ant's, just got lucky, but Big Lew has his in his 40% Carden with the same #'s as mine. I have never seen mine go into hold and 90% of the time the fails are less than 10. I have flown it out as far or farther than I flew it on 72mhz. I know this thread was on the new radio but hope my info helps.

Dom

[1bwana1]

Here are a couple of pictures of the installation. Any suggestions of insights will be appreciated.

Here you can see the Main RX sitting on the power expander, which is installed on the platform made by the two cross braces. This keeps the setup fairly high up in the center of the fuse. You can also see the first satellite RX installed on the opposite side of the fuse.

921main.jpg

here you can see the second satellite RX wires running back to where it is installed about 1/2 way to the tail and up at the top of the turtle deck.

921remote.jpg

[bodywerks]

It would be nice if the logger gave you some idea of what is considered acceptible, wouldn't it?
I do know that the transmitter sends several frames per second. it could be that you only lost one-one millionth of the total frames sent during the flight, which would be pretty darn good. Fades and frame losses are not as bad as holds. Getting that number to 0 is what I'd be focusing on.

[1bwana1]

Nick,
I agree, since they are giving us the data logger tool, they should give us an idea of what number ranges for each attribute is acceptable, and optimal. If I remember right the frame rate is like 20 milliseconds. If that is even close, there are millions of frames sent in 14 minutes, so 132 is probably an insignificant number. I have heard that in a good setup, a couple of hundred fades on each RX is pretty normal. Any holds indicate a problem. So I am thinking we should change the position and orientation on the RX's to get those numbers down.

Of course this info is not from JR directly. Once again, I thing they should give us some guidance on this subject.

[rede2fly]

great reading guys but, after reading all this i think i will wait for a year or 2 to see if most of the bugs are worked out. i would really be pissed to loose one of my planes due to lock out. but keep the info comming