Hi TheZoomer,

Just had a look at the Kionix KGF01 series gyros and they certainly look great. Love to get my hands on them. I wasn't even aware of their existance.
The big problem in the electronics industry these days is not being able to buy in small quantities.
Almost impossible, with 1000 or higher MOQs. To give you an example of how difficult it is: I work for an electronics component supply house and we are sole New Zealand distributors for Murata. I can't even get hold of Murata gyros in small quantities and we're the damn distributors here.

I don't mean to put you off your research. It was this never give up attitude that enabled me to find a source of supply for Tokin gyros in one or more quantities. Problem was their US$25 cost.

So ... yes .... if you can find a source for those Kionix KGF01 series in small quantities, I'd certainly be interested.

Cheers ADI

Hi ADI,

I THOUGHT you would like those. They are not the only guys producing this type unit. So maybe we can find a workable small quantity source.

I haven't been real aggressive as yet as I was unsure if you electronics guys thought we could integrate these and they were the TICKET. Hopefully you guys can intuitively conceptualize a simple integration. Am I correct, that these can simplify the board and programming??

Also, I am unsure of many of the optional parameters to chose, especially output. There are pulse and analog options out there. Looks like the vibration tolerance is good as well as the shock resistance.

I have since learned that "in the sensor biz", Piezio is considered to be very substandard technology for these purposes...may be part of the trouble.

I also found these.. [X(]by trying to find digital inclinometer sensors as used in digital levels. I plan to go find a digital level in a store and play with it a bit. Have to be careful though there are different setups...some more intertia sensitive than others. While pitch and roll gyros are good, they don't know where "back home" is, and normally need corrections for the various ways they wander.
http://www.frederickscom.com/product..._overview.html
These tilt sensors are available in variable viscosity so you could get a dampening effect to eliminate the sloshy jitters. I sat here playing with a glass of water, and boy it sure looks like something like a light oil would work great. Even with water, accelerations really have virtually no effect on the average surface plane. I almost think these may be able to eliminate gyros altogether...for the DF. For the moment, I can't see how, with these tilt sensors properly integrated, that a DF wouldn't practically auto hover. Maybe I have't thought it thru enough.

Here is something else to look at. Tilt sensor...with a mini slot car track built right into the board..phenolic must have been on sale??
http://www.atmos.com/product_details/at3510.htm

Boy, what jacks my crank is how stingy so many of these websites are with data. Jeez, with the web, if I was trying to sell something I would leave no word left out. Far as I know, server space hasn't killed any trees. And so many leave you clueless about how to get what they have.[:@]

Anyway..Cheers,
John

Hi TheZoomer,

>I sat here playing with a glass of water, and boy it sure looks like something like a light oil would work great?

Hmnnn ...... not a bad idea at all. A double ended paddle sitting on top of the oil with an axle through the middle. Coupling electronically to the axle's position relative to the glass would be simple.

Re tilt sensors (accelerometers): Problem with them is they react to acceleration in opposite phase to the tilt, thereby cancelling the signal. Just picture a small weight secured with a spring at each end. When you tilt it forward gravity pulls it forward, but the DF will also fly forward, and it's acceleration forward forces the weight backwards (cancels the original gravity pull). They're great for static tilt applications though. Better than a gyro.

Might pay to check out the acceleration effect on that glass of oil.

Cheers ADI

The zoomer,
you are correct about the gyro gain. they do absolutely nothing unless it is in the wind and when they get a big gust they begin to hunt but it keeps it level long enough that i can chop the throttle and land. Given that, why dont you guys just lower your cg and make your propellers booms a bit longer? any destabilization affect you may experiance would be severely dampened by the longer booms, the lower cg, and it would save you guys a lot of headaches and $$. that's essentially what i did, the gyros are just there for emergency purposes only. i dont know if my being able to control it is a matter of a stable design or the fact that i have been flying twitchy tailheavy overpowered aerobatic fixed wing aircraft for several years now, but either way, i am sure you guys could do it no problem. if it helps you could always tether it from the top with a light reel or something so that whatever you have it tied to picks up the slack, and you could learn to fly it that way without fearing a crash or something. that worked for me. the first flights i had were on a teather.

"Hmnnn ...... not a bad idea at all. A double ended paddle sitting on top of the oil with an axle through the middle. Coupling electronically to the axle's position relative to the glass would be simple."

Hi ADI,
Thanks for the compliment.
Went round and round with that last night. Seems, simplistically, that the DF must tilt FIRST to accelerate. If the initiation of that tilt can be terminated, instantly, by sensors, back to level...there would be no acceleration. Viscosity, physical sensor design, and sensor output handling would dictate the hunting frequency.
In a real helicopter, you can hover motionless while somewhat aggressively "stirring" the cyclic. While the control inputs, and results, are quite huge, you go absolutly nowhere, as they cancel out....if you do it correctly.

In the case of a desired acceleration...in an upright flying craft as the DF (sorry Clam..working on stability here,...not loops) the sensor would reestablish gravity relationship quite quickly, as prolonged acceleration isn't an issue. Also, rotorcraft can't pull "holeshots" so the tilt/acceleration argument should be quite minuscule.

What I thought might be quite peachy[sm=idea.gif] would be mounting the tilt sensors on servos....and having the tilt sensors directly command the ESCs. One problem there...throttle...haven't quite mentally gotten around that issue yet, but I see it possible, just need an overall modifier.

The reason for the servo/sensor thought was to easily overcome the arguments between the pilot and stability sensors. Also, Tx trim function could be utilized to zero out attitude, and help counter drift in a wind. It also seems a sensor commanded ESC is easier than one signaled by the Rx??

Did you get a good look at the glass vial sensors at:
http://www.frederickscom.com/product..._overview.html
The circular one looked real interesting???

The have a diagrams showing how they are constructed. No big deal, but they require AC supply current and the two signal legs appear to divide the input proportioned to tilt. I assume it's a restive function between the fluid and the electrodes.

BTW, How do they count posts here...and what is that contributor rank system ?? I show 16 for the last three at least.[sm=frown.gif][sm=frown.gif][sm=frown.gif]

Cheers,
John
"Box?......what box?"[sm=idea.gif]

you know, you can change your custom title. i think su;er contributor starts at like 150 or something That's how mine says "that is what clams are for!" but for a while i had "mondo contrubutor" and "overlord contrubutor" and even "administrator" for a while until the real admins tracked it down.
"

Clam,
As always, excellent advice IMHO. And you are so modest.
And loops...and more ( 8 pt rolls?) are good...just a bit stuck on this stability thing and trimming the $$$ out of this for now.

Maybe we could just adapt this simple method....[sm=idea.gif]

THE CAT AND DUCK METHOD OF INSTRUMENT FLIGHT.

Place a live cat on the cockpit floor. Because a cat always remains upright, he or she can be used in lieu of a needle and ball. Merely watch to see which way the cat leans to determine if a wing is low, and if so which one.

The Duck is used for the Instrument approach and landing. Due to the fact that any sensible duck will refuse to fly under instrument conditions, it is only necessary to hurl your duck out of the aircraft and follow it to the ground.

LIMITATIONS ON THE CAT AND DUCK PROCEDURE.

Get a wide awake cat. Most cats do not want to stand up at all. It may be necessary to carry a large dog in the cockpit to keep the cat at attention.

Make sure your cat is clean. Dirty cats will spend alll their time washing. Trying to follow a washing cat usually results in a tight snap roll followed by an inverted spin.

Use only old cats. Young cats have nine lives, but old cats that have used up their lives have as much to lose as you do and will be more dependable.

Beware of a cowardly duck. If the duck discovers that you are using the cat to stay upright, it will refuse to leave without the cat. Remember, ducks are no better in IFR than you are.

Be sure that the duck has good eyesight. Near-sighted ducks sometimes fail to recognise that they are on the guage and go flogging off into the nearest hill. Extremely near-sighted ducks will fail to recognise that they have been thrown from the aircraft and will descend to the ground in a sitting position. This maneuver is very difficult to follow in a fixed wing aircraft.

Use only land loving ducks. It is very discouraging to break out and find yourself on final for a lake or marsh, particularly if there are duck hunters around. Duck hunters suffer from temporary insanity while sitting in freezing water for hours and will shoot at anything that flies.

Choose your duck carefully. It is easy to confuse ducks with geese as many waterbirds look alike. While they are very competent instrument fliers, geese seldom want to go in the same direction as you. If your duck heads off for Cape York or Tasmania, you can be sure that you have chosen a goose.

Cheers,
John

[sm=confused.gif]
i have yet to try a loop or roll on this DF, but if i can get a saftey net it's going to happen sooner or later...
also, in order to use the cat and duck method on the draganflyer, it would require a lot of power to be able to carry a cat/duck payload let alone the fact that it would end up top heavy and the cat may stick it's nose into the props out of curiosity and you end up with a dead cat and a topheavy craft with no cat to observe your relative position to gravity with.
Cheers,
Clam

TheZoomer Wrote:
>What I thought might be quite peachy would be mounting the tilt sensors on servos....and having the tilt sensors directly command the ESCs. ........ etc....... It also seems a sensor commanded ESC is easier than one signaled by the Rx??>

This is how it's currently done. The sensors already directly control the esc/s.
The signal control coming from the RX is in addition to the sensors' control.


Cheers ADI

The Zoomer Wrote:
>...the DF must tilt FIRST to accelerate. If the initiation of that tilt can be terminated, instantly, by sensors, back to level...there would be no acceleration.>

"...the DF must tilt FIRST to accelerate" sorry ... don't agree on this one.
The eqilibrium of the DF is feather touch sensitive and can be upset easily, like trying to balance a ball bearing on top of a pin.
Even if you could employ all of NASA's resources to come up with a tilt signaling device with an accuracy of a ten billionth of a degree tilt error, it still wouldn't give you a perfect hover.

The DF may BE perfectly level with no tilt whatsoever and it may still wander off in one direction or another. The eqilibrium imbalance could be caused by a single prop geometry difference or motor power difference .....and any number of electronic reasons also. So perfectly level doesn't mean perfect hover. It just means "perfectly level, but heading off slowly towards mums bloomers on the washing line"

Tilt sensors just tell you when you're in the ballpark. Any acceleration info that can be sensed after this will then tell you that your perfectly level status is not what's required to maintain a hover. So we'd better change it to an imperfect level if we want to save mum's bloomers.

It's annoying, the fact that the laws of physics dictate that gravity and acceleration are like Siamese twins.

Cheers ADI

Off topic here with this one: Did you know that not a single dead animal has been found in Indonesia since the Tsunami struck ?
For example ... apparently all the elephants broke their chains or ripped the stakes from the earth and all headed for higher ground 5 minutes before it struck.
Source: BBC World Television News

Cheers ADI

just goes to show that we are not part of the intelligent life bracket.
Cheers,
Clam

Hello all you hover dudes,

I want to introduce myself and ask a question. I have loved reading this thread on this forum. I am yet another guy who about a year ago set out to make my very own DF'ish type of flyer but it's still sitting nice and still on the old workbench. Hopefully I can get it airborne sometime soon though, and then I too will be in the club of those who can post their picture of a homemade, flying machine. I can't wait. All of you who have broken the flight barrier, good job. Man I am sure impressed! My DF solution consists (presently) of the micro-p flavor, I am using a philips 8051. I liked it because the PWM outputs were so easy, and I had it and could program it. I was attempting to use a tilt-sensor, (ADXL-202) but knew all along in the back of my mind that there had to be a catch. I mean I could get those samples for free from AD, so that was a kiss of death right upfront. After reading posts on this forum, I finally realized more detail about it...

Anyway, my question is about piezo gyros, and I apologize if this has already been addressed. (I haven't seen it though) I am wondering if there has been any evidence or other feelings about which ones to use? I know that is kind of hard, because it might also depend on which kind you can even get. I saw that ADI had located a source of TOKIN gyros for $25, and I have now found a source of TOKIN for about $27, and I have found a source for Murata at $30 each. (These are for small quantities). Does one perform better than the other? I mean if you can actually choose, why not. At that price, I hate to buy the ucky ones. Also, has anyone looked into the Analog Devices gyros? I thought someone said you could get samples, but today more and more electronic mfgs are charging you for "engineering samples". The AD stuff looked cool to me except for trying to mount that dang BGA... sheesh. A Murata rep did tell me that if you want to buy the 1500 pieces, then your price is more in the $10-15 range. He also said Murata is quite picky who they will sell them to, because they don't want to sell them to somebody who is using their parts in a fashion that Murata doesn't approve of. If someone here had a buddy who worked at a camera/camcorder manufacturer I bet they could go raid a parts bin and still make money selling to all of us poor saps who only want a handful.

Your comments would be appreciated...

Thanks all!

Welcome A2,

You should have dropped in and said hello way back. We love photos here, so if you've got some of your DF then let's have em.
Re gyros: AD have scrapped their free sample program. Samples can only be procured through your local AD distributor. BGA devices ? Forget it ! Even with the correct IR ovens, you have to xray the pcb to see if it soldered properly. TSSOP packages are trouble enough.
Haven't compared Tokin and Murata gyros on the workbench, so can't answer your 'which is best' question. I'm interested to know your source of supply for both brands though. This is the one part of the DF that is hard to get hold of.
Strange thing about the Japanese semi manufacturers ..... as you say .. they're very fussy who they supply to.
There prescence on the web is limited and application info for their product is scarce.
It's a damn cultural thing I'm told. Look at the western companies .... TI, Maxim and National for example .... they can't wait to give you samples in order that you'll advertise their product for them.
Geez, those 3 companies are the reason my DF can fly !

Re Phillips 8051 ..... does this have multiple output compare registers ?

Cheers ADI

Yes, I should have said "hey" alot sooner than I did. doh!

The exact part I am using from philips is the 89C51RD2. I just happened to use it for other projects at work, so they are plentiful and I have several boards to fool around with them. It has 5 16-bit compare modules, (I never did figure out what to do with the 5th one) but I only ran them in 8 bit mode. I couldn't see needing any more than 8 bit resolution on the motors, and as I fooled around, it could amply control the motors. I am using an (2nd) external crystal on the micro to generate the PWM freqs, which I found from goofing around worked good at around 1 khz. I got 10 khz for free from the micro, and it worked, but the resolution was not good. The micro had no problem reading the radio info, and I have it decoding the tilt sensor as well. At that point, I started looking into making it go into the air, but haven't finished building the frame. The tilt sensor was nice, because it has digital outs and I didn't have to A/D anything. (That's where this micro comes up short, no A/D...) So, my next step is to choose a piezo, and add the A/D junk. Dang, I wish gravity wasn't an acceleration, because then I wouldn't have to beg for parts.
The sources for the Tokin gyros I got quotes on are from a rep in CA, called Bravo Electro components. The murata source is from a rep we use here in Utah. Both told me I could order any quanitity. that's why I was so happy to get firm pricing, because they deal in selling parts. I was told by the murata rep that the one reason the cost is so high is that the one-offs come from "sample stock" and they have always been able to get them for $30, take it or leave it. Yes, the western companies (Maxim is a prime example) don't even bat an eye asking for a couple of freebie parts. That's what's so frustrating from murata, I have got samples of their ultrasonic transducers before, I figured a stinkin piezo would be cake. silly me.

Hi A2,

You're lucky to have so many output compare regs.
Not so lucky with my AT90S2313 Atmel, so I've had to use the age old 'triangle osc + 4 comparators' PWM trick. Works well though, when fed with a 4 channel D/A convertor.
There are lots of 4 channel A/D chips around if you do try some gyro chips.
You'll shelve your ADXL202s once you try them.

Cheers ADI

Well guys,
Once again i am looking into the CSM heaidng lock gyros. While mine flies ok with no gyro imput for the most part, it still requires A LOT of stick correction constantly, and would like to have a relatively stable hover so that i can do more of what i want to do rather than worry about focusing all my attention just keeping it still. I am going to try and sell my pg 03s and use some of the extra $$ that i am earning by painting game consoles. I believe a heading lock is the way to go with this becuase it will sense those minor things and it will continue to correct until it gets back on track rather than jsut a quick blip. Also, i am looking into longer prop shafts to lower my CG. i will let you know of my progress.
Cheers,
Clam

Hi Space,

Personally, I wouldn't increase the prop/gearbox shaft at all. In fact I'd do the opposite in order to keep the prop as close to the top bearing as possible. In theory, the longer the shaft, the more preasure you'd put on the bearings, especially if the shaft was not perfectly straight.
Much easier to remake the frame with a pair of V shapes, in order to lower cg. CF rods glue great with 24 hour epoxy.

PS: How much you selling your PG-03s for ?

Cheers ADI

the problem with that is that the body currently is made of styrofoam with a balsa bottom which acts as a landing pad, so it would not support a landing structure such as that. remember, a lot of my landings are less than perfect so between the styrofoam being flexible, durable and light and the balsa keeping it from crushing, it's really the best landing pad. a CF stucture on the bottom would never stand up to a 2 lb craft landing at a 60 degree angle at above optimal landing speeds. Correction, the CF would stand it but would break the bottom of the craft. I dont really care if it puts more pressure on the top bushing. I can replace them with ball bearings for about $2.50 a pop and if lowering my CG relative to the prop position costs me $12 every couple of years i would say it is better than constantly repairing it every time i have a hard landing. I am not going to sell my pg03s until i can get my hands on the heading locks. i assume that you are looking for the actual gyro chips? I will probably sell them for fair market value, due to the fact that cash is tight, and not only am i trying to get two heading lock gyros but i am also trying to save up for a 3d electric foamy to practice my 3D stuff on, and for a good brushless set up it's going to cost me about $350. Although, If i am unable to get a retailer to get me some kind of a discount between them, i probably will not sell until i can afford the gyros normally in which case i would probably sell them to you for a bit less becuase $$ is not as tight. we will see. Either way, we should start up a website containing all our efforts and ideas, stories of our successes and the journey getting there, pictures, videos etc. What do you guys think? I only have 5 megabytes, but if somebody were willing to host it i would gladly contribute. That way, rather than having people search through this thread to see what we are doing, we could jsut post links to the website so people can referance info and stuff, but still use this thread to communicate, and maybe the website will grow. What do you guys think?
Cheers,
Clam

Hi Spaceclam and All,

With a bit more web sleuthing I found out that the Futaba GY series gyros
http://www.towerhobbies.com/products/futaba/gyros.html
utilize the micromachined mims chip gyro technoligy.

I think they all come with a normal, and heading lock option select.

To give an idea of the performance of this gyro chip, it is used in the new I bot wheelchair that elevates the rider to face to face height..WHILE BALANCING ON TWO WHEELS!!


And here is where the actual gyro component comes from:
http://www.spp.co.jp/sssj/silicon-e.html

Any ideas on how to acquire some of these??

Cheers,
John

TheZoomer,

Only those wheelchair guys don't tell you about the accelerometer chips that are Kalman filtered with those gyros to obtain a low drift angle. Check out some of the balancing robot projects online and you will see what i'm talking about.

Take a look at this one: http://www.tlb.org/scooter.html

Hardyus.

Hi All,

Hardyus..... How are you getting on with your DF electronics ?

Spaceclam ... a website sounds like a pretty cool idea. What will we do though when the site becomes famous and we have 5000 people worldwide all wanting to build DFs ? he he

Cheers ADI

"Only those wheelchair guys don't tell you about the accelerometer chips that are Kalman filtered with those gyros to obtain a low drift angle. Check out some of the balancing robot projects online and you will see what i'm talking about.
Take a look at this one: http://www.tlb.org/scooter.html "

Hi Hardyus,
Thanks, Yep had been at http://www.tlb.org/scooter.html a while back. It seems that the wheelchair source in the pic is not associated with the "scooter" source. Also, Johnson & Johnson is the parent company of that wheel chair project.
I found the wheel chair, and the Futaba link posted in the applications links on the gyro chip site http://www.spp.co.jp/sssj/silicon-e.html indicating the chips come from them.

What I am really looking for is a good board design, component list, and programming for dummies....seems like the prices at http://www.tlb.org/scooter.html are a quite high ( 1 axis $150 )? for a DIY project?? and I can't see what sensors they are using.

It also appears that the chip from http://www.spp.co.jp/sssj/silicon-e.html outputs 3 axis from the one chip???

Update..
I e mailed them and received this response:

Dear Customer

Thank you for your interest in our angular rate sensors.

Because of your geographical location of your enquiry, this will be handled
by our UK office:

Silicon Sensing Systems Ltd.
c/o BAE SYSTEMS PLC
Clittaford Road
Southway
Plymouth
Devon
PL6 6DE
England

Attn: Sales

tel +44 1752 695695
fax +44 1752 723331

e-mail: [email protected]

Your e-mail has been forwarded to the above party.
If you have problems contacting them, please advise me.

With best regards,

Chuck Nakamura
general manager - sales
Silicon Sensing Systems Japan, Ltd.
(c/o Sumitomo Precision - gyro production site)
1-10 Fuso-cho, Amagasaki, Hyogo 660-0891
Japan
tel +81 6 6489 5868
fax +81 6 6489 5910
email: [email protected]
www.spp.co.jp/sssj/

No results as yet from the UK.

Those with interest, and with better credibility, than I, are welcome to follow this up. I am quite ignorant in the lingo of this field. In these component sourcing efforts perhaps we could get a leg up by forming a "professional project group" image??

Cheers,
John

Hi Spaceclam & ADI,

About the website...I have MS front page and.. I THINK.. I am a bit handy with it (I created this RE site and all the graphics from scratch, for my friend Lynda: http://www.lyndamelnick.com/ ) This domain is being discontinued soon, so look quick.

I wouldn't be able to host, or be the primary webmaster, but I could build a nice core site and transfer the files... and pinch hit updates when I am available.

Guys take a look at THIS DF. This guy is going full bore on fully autonomous flight. This is just a side trip.
http://www.nongnu.org/paparazzi/images/cam4_bench.jpg

Cheers,
John

ADI,

Made my PCB last weekend, positive presensitized board, it all came out beautifully, except after drilling i found out that some of the holes were way too small I redesigned it, so now i just have to do it all over again.

TheZoomer,

Sorry for the confusion, i didn't mean that the gyro source was the same for the scooter & wheelchair. What I was trying to blurt out was that the chair is so stable because they are probably using accelerometers to compliment the gyros, and to minimize the gyro drift. As Danvel pointed it out in an earlier post that with gyros only if you start out hovering with center stick, it will soon drift away from there and you'll have to correct with the stick, and even then you won't get a stable hover. Kalman filters seem to correct for gyro drift, and that would probably make it pretty damn stable. I've seen that wheelchair in Disney's Epcot center, it stays balanced upright like that for a long time.

Hardyus