Canard Community Forum  

Go Back   Canard Community Forum > Instuments / Avionics / Electrical System
FAQ Members List Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Display Modes
  #1  
Old 05-13-2007, 01:10 AM
rviglierchio rviglierchio is offline
Member
 
Join Date: Apr 2006
Location: Davis, CA
Posts: 72
Default Does anyone have one of these or want to build one?

I would love one of these to go with my pitch trim.
A friend that builds a lot of electrical things says it's as easy to build several as it is to just do one, and the price of the boards goes down with quantity. The fellow that designed it quit making them.

http://hometown.aol.com/ccady/eztrim.htm

From his web site:

EZ-Trim
Altitude Hold Unit

Description & Specifications

What is it?

The Ez-Trim is a small low cost unit that provides basic altitude hold function for homebuilts. I wanted a economical unit for my E-Racer canard. I could not find one, so I started designing my own. After several months and over half a dozen different test units, I had a unit that worked. It has been used and tested with a MAC pitch trim servo attached to the PTH trim lever and the Strong pitch trim servo. It will maintain the selected altitude very well in smooth air conditons. In turbulence it will keep trying but you may prefer to fly manually. The key is getting the unit adjusted to the plane properly using the response adjustments provided. It has front panel adjustment screws to allow you to adjust its response to your particular homebuilt.

It is intended to provide missing altitude hold when used with a wing leveler like the Navaid Devices unit. It sends control to the relays that control the electric pitch trim servo. The unit is designed to provide altitude hold function using a pitch trim servo through relays. The unit does not have the current capacity to directly drive the servo.

Since the unit controls using the pitch trim it is easy to override if there is a malfunction.
Reply With Quote
  #2  
Old 05-15-2007, 08:39 AM
WileEZ's Avatar
WileEZ WileEZ is offline
No Freaking Clue!
 
Join Date: Apr 2007
Location: Kirkland, WA
Posts: 56
Default

Looks interesting, I bookmarked this page for later review.

Thanks for letting us know about this!
Reply With Quote
  #3  
Old 05-15-2007, 11:41 AM
dpaton's Avatar
dpaton dpaton is offline
Future Cozy Driver
 
Join Date: Oct 2005
Location: Streamwood, Iiilnois
Posts: 259
Default

That looks deviously simple. I've actually got all the parts to build one within arm's reach here at work, except instead of the little Basic Stamp, I'd use an Atmel and some C. It's a slick product, no doubt, but from the development page, it sounds like he needed a little help with the analog/power side of things.

Small time manufacturing things like that will absolutely kill you. There's a reason I'm not actually going to build one
__________________
This is not a sig. This is a duck. Quack.
Reply With Quote
  #4  
Old 05-15-2007, 12:46 PM
Wayne Hicks Wayne Hicks is offline
Member
 
Join Date: May 2004
Location: Carrollton, VA
Posts: 1,376
Default

The unit debuted long ago, certainly before I came on board (circa 1998). I don't think it received a wide acceptance. I don't see it on many of the instrument panels. The unit was in the Cozy III that I'm flying. We took it out because it didn't work so well.

I'd get other peoples' opinion, good or bad.
__________________
=============
Wayne Hicks, Cozy Plans #678
http://www.maddyhome.com/canardpages...cks/index.html
Reply With Quote
  #5  
Old 05-15-2007, 03:19 PM
rviglierchio rviglierchio is offline
Member
 
Join Date: Apr 2006
Location: Davis, CA
Posts: 72
Default Improvements

The fellow that built it for his E-Racer kept improving it until it works pretty well according to him. I'm going to build a couple and see. I'll post the results. Thanks for the input Wayne.
Reply With Quote
  #6  
Old 05-15-2007, 06:17 PM
Nathan Gifford Nathan Gifford is offline
Nathan Gifford
 
Join Date: Oct 2004
Location: Tickfaw, LA
Posts: 897
Default

I think the original problems with the unit were the way it controlled. At least from reading around the web that it lack PID control. Apparently later versions took care of that problem.

The other major issue is whether you really want to control your altitude with your pitch trim servo. If the pitch trim system were to fail you might have to abort your flight depending on where in the servo's travel it failed.

Another problem fot this type of control is the slowness of pitch trim. This can create a lag, for good or bad, that may make the system hunt more under certain conditions.

I believe if you read around about it the flyers that were most dissatisfied with it probably had early versions. Later versions with better control algorithms reportedly did better.

Still it really looks like if you want good altitude control you need a pitch servo and not a trim servo.
__________________
Nathan Gifford
Tickfaw, LA USA
Cozy Mk IV Plans Set 1330
Better Still --> Chapter 9
Reply With Quote
  #7  
Old 05-16-2007, 01:28 AM
rviglierchio rviglierchio is offline
Member
 
Join Date: Apr 2006
Location: Davis, CA
Posts: 72
Default PID?

Nathan,
Can you enlighten me as to what PID means? Am new to this stuff and plead ignorance....
Reply With Quote
  #8  
Old 05-16-2007, 02:50 AM
chasingmars chasingmars is offline
ChasingMars
 
Join Date: Apr 2006
Location: Ottawa, Ontario, Canada
Posts: 418
Default

Quote:
Originally Posted by rviglierchio View Post
Nathan,
Can you enlighten me as to what PID means? Am new to this stuff and plead ignorance....
Proportional-Integral-Derivative...

In automatic control, basically, you want to have an actual physical parameter (lets say, your altitude) match a set value, which could be static (altitude hold), or changing (a descent profile for example)

PID control is an effective and relatively strightforward method of generating the forcing function, or more simply a controlled output such as, in this case, elevator deflection, to make the system meet the desired constraint.

So Proportional control is the easiest to grasp. The system dials in, lets say, 1 degree of elevator deflection for every 100 feet off altitude you are. This 1 degree/100 feet is an expression of the controller's Gain. Proportional control has the gain proportional to the measured error, essentially, in a direction to correct it, i.e. the output signal is the proportional gain multiplied by the measured error.

A note, relevent to the example, you usually need limits outside of which the system can't be expected to be stable, for example, if you were 5000 feet off altitude, trying to command 50 degrees elevator wouldn't help, you need to limit total deflection to keep within the flight envelope (i.e. limit the up elevator to keep you away from stalling and the down to keep you safely away from Vne (and hopefully, terrain)).

Integral control also produces a output, but as a function of time as well as measured error. Basically, it integrates the error over time instead of just taking the measured error directly. The integral component of the output signal is then a result of this integrated error multiplied by the integral Gain. This lets you reach your end value more quickly as the error sort of "builds up" over time.

Usually, you'll have a P-only, or a PI or PD or full PID controller (I've never heard of a ID controller, but it's possible I suppose), this just means there's a component of that nature, and the actual output is the sum of the component outputs (at least at a basic level, but it serves for a lot of circumstances, this stuff can get as complicated as you want it to, but the simple approachs work remarkably well in most situations).

Differential control sort of damps the system, to try and reduce overshoot. Basically this adds a component based on the rate of change of the error, again multiplied by a gain, if the error is changing quickly, then, there will be an added component to the output signal. To be stable, this gain would be negative, so that if for example, you've reached your altitude, but you're still climbing through it, there's a component of the forcing function working against that rate of change, right then (and indeed, in a well designed system, before then), rather than you having to actually exceed the altitude before there's some forcing function pushing you back to the desired point. This is important in reducing oscillation in the control system.

The real trick of course, is choosing your Gains. Choosing a bad set makes the system unstable, and since there are three degrees of freedom in a PID controller, tuning the system can be non-trivial if you go by trial and error (not to mention, potentially dangerous). How to chose them intelligently isn't trivial, but there are established methods. Most of these rely on knowing how the system responds to changes over time, i.e. the actual flight dynamics of the aircraft in this case.

There you go. I knew that postgrad work in robotics would have a use someday!

(Edit... after posting realized you asked Nathan, rather than as an open question... didn't mean to be pedantic, but hope this serves anyhow)
__________________
Craig K.
Cozy IV Plans #1457 - Building Chapter Six!
http://www.maddyhome.com/canardpages/pages/chasingmars/index.html

Last edited by chasingmars : 05-16-2007 at 03:11 AM.
Reply With Quote
  #9  
Old 05-16-2007, 09:31 AM
Nathan Gifford Nathan Gifford is offline
Nathan Gifford
 
Join Date: Oct 2004
Location: Tickfaw, LA
Posts: 897
Default

Well besides what Craig has said, you also need to realize that using the pitch trim servo also adds its own lag. This certainly makes tuning a PID controller all the more difficult.

I sure the big problem was putting enough smarts in the program to handle both smooth and rough air.
__________________
Nathan Gifford
Tickfaw, LA USA
Cozy Mk IV Plans Set 1330
Better Still --> Chapter 9
Reply With Quote
  #10  
Old 05-16-2007, 11:56 AM
rviglierchio rviglierchio is offline
Member
 
Join Date: Apr 2006
Location: Davis, CA
Posts: 72
Default This, from the designer....

"My unit does the corrections and you can adjust it, but where it comes short is where you have rough air or large disruptions. The latest version of the software that I tried used an adaptive control called PID that worked better by adjusting the response better on the fly.

My unit it totally controlled by the software which I was improving as I tested it. The Basic Stamp II cpu has a very limited memory for the software. Luckily the Basic Stamp II is so popular that there are are other cpu's that plug into the same socket.

It was fun to try to come up with a unit that I could work on and didn't cost a bunch of money."

Sounds like he may have solved some of the problems. The parts are less than $200 so I'll try one and get back to the forums.
Thanks for the input.

PS: Wayne, what type of box or enclosure was your in?
DV
Reply With Quote
  #11  
Old 05-16-2007, 12:21 PM
Nathan Gifford Nathan Gifford is offline
Nathan Gifford
 
Join Date: Oct 2004
Location: Tickfaw, LA
Posts: 897
Default

Its real charm is the low price of the gear. Since I think what you are using is the translation of the last comments about EZtrim, I think rough air is still a problem but not as much as it was before.

If you have the EZtrim source code, it sure would be nice. STAMPs are very nice devices but there are better but not as easy to program. One of the things you might want to add to EZtrim is code that warns when the system is doing too much hunting and when it is drift outside the target altitude.
__________________
Nathan Gifford
Tickfaw, LA USA
Cozy Mk IV Plans Set 1330
Better Still --> Chapter 9
Reply With Quote
  #12  
Old 05-16-2007, 01:27 PM
swinn swinn is offline
Member
 
Join Date: Jan 2005
Location: San Diego, CA
Posts: 76
Default

This is an interesting thread. Has anyone ever used a hobby servo to drive the pitch axis of an altitude hold? Some of the new 'Digital' servos provide a lot of torque in a very small, light package. The heftiest hitec servo provides 333 oz/in of torque at 6V, they $109.

HiTec HS-5955TG Specifications

Of course you would need to come up with an external method of connect/disconnect from the elevator, and it would have to be fool proof. Driving the servo from a microcontroller is a pretty trivial task as they accept a simple PWM signal. Even a Basic Stamp can do it.

--Scott
__________________
--Scott
LongEZ N1LZ
Photos
Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 04:16 PM.


Powered by vBulletin® Version 3.6.7
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.