42 replies to this topic

[Mr.RC-Cam]

Hmm, I cant seem to veiw these but I see he is using transistors at the sensor end, is that right ?

It looks to me like he basically copied Maynard's circuit, but used more modern parts. FETS are in the front end.

[Delcam]

as far as I can read chinese...he uses the gate of the fet as the sensor??

Hans Delemarre

[Mr.RC-Cam]

It all looks like greek to me. But, I think the basis of his version is that he did not use the polonium ionization doo-dads. Instead, it appears he used a FET at each wingtip to act as the sensor. There is also a mention of a active filter stage, gyro, and DSP back end.

[Terry]

GYRO !?!
That makes me feel it did not work...

Terry

[dleroi]

Anyone care to revive this topic? I've done a little "translating" and have sort of figured out what he's doing. BTW, the gyro wasn't part of the project. Rather, it was used to verify the output of the thing.

- Don

[Terry]

Yes if anyone has any practical experience it would be interesting. I for one would like more info on the fet sensors.

Terry

[Mr.RC-Cam]

Yes, it would be nice to see someone try it out. If you have translated it, then it would be good to see the white paper published in English.

[dleroi]

Terry and Mr. RC-Cam,

I didn't mean to imply that I had translated the whole document. I used bablefish.altavista.com to translate certain phrases and sentences. The translations read a lot like some of the users manuals we've all seen come out of Asia, but it was sufficient to give the flavor of what he was doing.

I've adapted his circuits to use components I had on hand, or could get my hands on, and I've breadboarded a bit of it to see if it shows any promise. The FETs, with 1" antennas, are amazingly sensitive. They pick up my waving a piece of plastic from across the room. Of course, that's what we want to filter out!

I'll know much more within the next few days and will post my findings here.

- Don

[Terry]

I have not done done any testing myself but I would think the layout and the material its buit on are critical. A charge must not be allowed to build up around the sensors, the response time has to be zero.

Please keep us informed of you results Don, I will have a go myself for sure if you can make it work.

Terry

[dleroi]

Is either of you familiar with Helmut Lelke's use of this method in his gliders? It appears he's been doing it successfully for over 20 years. I came across an article in "RC Soaring Digest" where he talks about Maynard Hill helping him with his initial attemps. Interesting read. Let me know if you haven't already seen it and I'll try posting it.

- Don

[Temporary Insanity]

Helmut and I had a pleasant discourse about the system and I came to some conclusions which I I'll share here:

1. Overall, I like this system and almost used it but chose in favor to the pyrometer (CoPilot type) instead

2. The isotope has a pretty short half-life of 170 days. This means semi-frequent replacement of the isotope. If not replaced the system tends to act more and more like a gyro instead of a wing leveler.

3. I thought it would be a better system if the long, high impedance coaxial lines were replaced with local, high quality opamps right at the sensor. These are cheap and available these days.

I hope this helps!

[dleroi]

TI,

Thanks for the info. I've been trying to contact him to no avail. His email address on CRRC doesn't work.

I agree with you on placing the FET stages at the extremities and that's what I've done in my experiments. My hope, though, is to build something that doesn't require the isotope.

It may be wishful thinking, but I don't see any sign of isotopes on his Albatross II photos or drawings. Were your discussions with him pre, or post, Albatross II?

- Don

[dleroi]

I've decided this is not a viable alternative for stabilization - at least not for me. I built a clone of the Chinese fella's system. It does actually respond to changes in tilt, but as TI suggested, it acts like a gyro. Maybe with the polonium it would behave as advertised, but that isn't the way I want to go. I'll probably attempt the accelerometer/gyro/Kalman filter solution, next. Meanwhile, I have my co-pilot and I'm attempting to gain an understanding of Kalman filters.

- Don

[Mr.RC-Cam]

That is sad to hear. Is there any chance the gyro-like behavior subsides at flying altitude? That is to say, does the proximity of the ground during bench testing create the problem? Or, is this just the way it is regardless of setup?

[dleroi]

Mr. RC-Cam,

I suppose it could be "ground effect", but the behavior is the same as T.I. says happens when the polonium weakens. I was hoping that the polonium was originally used to increase the sensitivity and that it might not be required with state-of-the-art circuitry. I've sent a PM to Helmut Lelke, asking him to describe his current sensors, but haven't received a reply. If I were to find out that he no longer uses polonium, then I could get excited about this, again.

- Don

[dleroi]

BTW, Polonium-210 is purported to have been the substance used to kill Alexander Litvinenko, the former Russian spy, last week!

- Don

[Terry]

Ahh..maybe he was working on electrostatic stabilisation !

Terry

[macboffin]

Did a bit of research on this system at London U, details lost in the mists of time, (I've been flying since Pontius was a Pilot ha ha!) but do clearly remember that one disconcerting thing was that the Eath's static field has bumps over trees etc, (let alone hills) and also varies over water. We were looking into using it for terrain following for a mine detecting UAV. Gave up because it was too fiddly and twitchy!

[Terry]

Are you saying it can not be relied on ?

Terry

[macboffin]

Definitely not at low levels. Higher you go the better. The "bumps" in the field are like air flow, the further from the obstacle, the less "bump" gradient. Imagine a small hovercraft flying over "Sleepng policeman" speed bumps at various heights, you can imagine the effects. Flying alongside hills/ridges the 'plane wants to turn away , the sensor nearest the hill thinks it is higher than the other sensor. Very sensitive, two inches over a seven foot span gets a response. Biggest problem, having to recalibrate every time out, since the rate of voltage-to-height changes with weather etc. No doubt adequate computer power and suitable logs onboard could make it workable, but there are easier ways!