Matronics Email Lists :: View topic - Open source product development for OBAM aircraft

Matronics Email Lists
Web Forum Interface to the Matronics Email Lists

Get Email Distribution Too! FAQ Search Memberlist Usergroups Register

Profile Log in to check your private messages Log in

Open source product development for OBAM aircraft

Matronics Email Lists Forum Index -> AeroElectric-List

View previous topic :: View next topic

Author Message

nuckolls.bob(at)aeroelect
Guest

Posted: Mon Apr 09, 2012 11:20 am Post subject: Open source product development for OBAM aircraft

At 11:57 AM 4/9/2012, you wrote:

 	  Quote: 	
		  



To that goal, I'd like to suggest using an Arduino.


	


  This is a highly capable device but a bit of an

  overkill for things like wig-wag flashers, voltage

  monitors, etc.



  The challenge is to exploit the scope of

  projects that can be implemented in a user

  friendly package of the smallest practical

  size, cost and parts count. An already stuffed

  Arduino board has no aviation friendly i/o

  or even an enclosure for $30. I would expect

  the wig-wag flasher to be drop-in ready for

  use on an airplane and have total bill of

  materials under $20 or so.

   Bob . . .

enginerdy(at)gmail.com
Guest

Posted: Mon Apr 09, 2012 11:42 am Post subject: Open source product development for OBAM aircraft

That's true, but it's fairly simple to 'roll your own' Arduino with just an ATMega and a particular serial and reset schematic. I haven't done the design exercise, but an ATMega168 would probably do the job, and they're as little as $2.31 on Mouser.

The FTDI serial chips are only $2 nowadays too, apparently, if you're inclined to make it USB programmable.

There are definitely other solutions, but I've had a surprisingly good experience with these things.

Daniel

On Apr 9, 2012, at 2:19 PM, "Robert L. Nuckolls, III" <nuckolls.bob(at)aeroelectric.com> wrote:

 	  Quote: 	

 At 11:57 AM 4/9/2012, you wrote:

> 

> 

> To that goal, I'd like to suggest using an Arduino.

 This is a highly capable device but a bit of an

 overkill for things like wig-wag flashers, voltage

 monitors, etc.

 The challenge is to exploit the scope of

 projects that can be implemented in a user

 friendly package of the smallest practical

 size, cost and parts count. An already stuffed

 Arduino board has no aviation friendly i/o

 or even an enclosure for $30. I would expect

 the wig-wag flasher to be drop-in ready for

 use on an airplane and have total bill of

 materials under $20 or so.

nuckolls.bob(at)aeroelect
Guest

Posted: Mon Apr 09, 2012 6:41 pm Post subject: Open source product development for OBAM aircraft

At 06:45 PM 4/9/2012, you wrote:

  	  Quote: 	
		  --> AeroElectric-List message posted by: "gregmchugh" <gregmchugh(at)aol.com>



 Bob,



 One question I forgot to ask in the previous note. When I looked at

 using the Perihelion Wig-Wag for my LED landing lights I was

 surprised at the compact size and the capability to switch the loads

 on standard landing lights. Is there something unique about this

 design to handle the 250 watt load? 	


   You'd have to check with Eric as to his

   design philosophy but based on what is

   generally known, getting the heat in the

   FETs under control has two components.



   (1) Thermal resistance from the silicon to the

   ambient. Like ohms in current flow, thermal resistance

   has a quotient most often stated in degrees C in rise

   per watt of power in the junction.



   (2) Keeping power dissipated at a minimum by selecting

   a FET with the lowest practical on-resistance in

   combination with a healthy overhead for rated voltage.



   Check out this array of 30 volt FETs.



  http://tinyurl.com/6vxhkxg



   Then let's select a 2.1 milliohm resistance at

   4.5 volts of gate drive. Nice specs but it's

   a bare chip intended for installation by bonding

   into a hybrid device. Okay back up let's leave

   the resistance alone and select 20 volts where

   it says we'll find 153 devices to select from.



   Lets go over in the array and look at surface mount

   parts with leads (67 choices).



   The next thing we're interested in is size. Just

   for grins, hit the SO-8 package (34 choices).



   Notice that we have three power dissipation choices.

   Let's go peek at the single 3.5 watt device.



   That gives us the IRF7811 with an logic level

   drive on resistance of 14 milliohms. Now, with

   the proposed 45 watt devices, let's assume 5A

   max at minimum bus voltage so 5A(squared) x .014 ohms

   equals 350 milliwatts. Hmmm . . . not bad.

   

   Let's go get the detailed data at:



  http://tinyurl.com/6qpt9cf



   We see there that thermal resistance junction-to-

   leads is on the order of 20 degrees C per watt.

   .35 watts will give us a 7 degree C rise on the

   junction over the temperature of the board. Now

   here's where it gets a little sticky. Our board

   is epoxy glass with most of the copper removed

   and the whole thing is enclosed in a plastic case

   with no ventilation. What we don't know is the

   thermal resistance board to ambient wile boxed up

   in this case.



   I'll go measure it in the next few days. This is

   done by cutting a piece of copper clad the same

   footprint as the proposed board. Glue a power

   resistor and thermocouple to the board. Close

   it up in the plastic case. Dump some handy amount

   of power into the resistor, say 1 watt. Then see

   what the temperature reads after an hour or so.



   This will yield the best possible heat transfer

   situation for one of these FETS soldered to the

   board. I'm thinking the 7811 might do the job.

   They're stock at Digikey for 1.50 in small quantities.



   Since this is a plastic enclosure, we'll qualify

   to cabin environments so 40 to 50 C is about the upper

   test limits.



   Okay, just for grins let's see how a pair of SO-8

   chips goes into the space previously occupied by

   three DIP4 chips and some do-lots components.



 [img]cid:.0[/img]



   Well, lookee there. Not only is there plenty

   of room, we might even put one on top and the

   other on the bottom so as to leave room for leaving

   more copper on the board around the chip.



   This exercise in cut and try on a four-cornered

   envelope with the final confirmation or denial

   of choice being a full power test in the temperature

   chamber. I have one of those.



  http://tinyurl.com/7yrrhhq



   So fortunately, we have access to quite a bit

   of experience to drive the corners of the envelope

   toward the component of choice.



   I'm pretty sure we can get a 5A switcher in

   this enclosure with comfortable head room

   for power handling. So now your schematic

   looks something like this:

 [img]cid:.1[/img]



 

   When we went from do-lots to wig-wag, it got a lot

   simpler.  Two outputs driving power FETs, two inputs

   to control for three states of function (OFF/WW/ON).



   So, first cut at the design looks good. If you can make

   the silicon sing, dance and do dishes, I'd say we're

   50% of the way to having a useful product that meets

   design goals and offers a user friendly envelope

   and wiring interface.



   Bob . . .

14cabde3.jpg

Description:

Filesize:

111.6 KB

Viewed:

14302 Time(s)

14cabe31.jpg

Description:

Filesize:

70.35 KB

Viewed:

14302 Time(s)

nuckolls.bob(at)aeroelect
Guest

Posted: Mon Apr 09, 2012 6:50 pm Post subject: Open source product development for OBAM aircraft

 	  Quote: 	
		  

Enough options to satisfy most of the potential users...


	


   Exactly. For years I tried to find a crack into

   which I could drive a wedge for distributed processing.

   We had dozens of little black boxes for current sensing,

   time delays, logic, thermostats, etc that all used discrete

   components and were designed 30 to 40 years ago.



   I even demonstrated that I could make huge reductions in

   parts count with similar reductions in size and weight,

   plus take advantage of pick-n-place assembly. But as soon

   as I mentioned software, everyone turns white and runs

   out the door.



   It didn't matter how simple the software task might

   be, those-who-know-more-about-airplanes-than-we-do-

   in-Washington insisted on piles of paper and buckets

   of holy-water with DO-178 'certification' and configuration

   management.  A device with $20 worth of parts to do

   some simple task could cost $100,000 to qualify. So

   much for modern marvels.



   Hence, anything software was farmed out in the form

   of "do-everything" flight and airframe management computers

   with a $10,000 bill of materials and a 10 million

   dollar cert program. What's more, when it didn't work

   right, we had somebody else to throw rocks at and then

   wait around with everyone camping on a work order.



   Now, you and I have a chance to show the guys at

   Beech/Raytheon/HBC how to do a responsible job of

   incorporating software with a distributed processing

   approach to black box development.



   Ought to be fun.

   Bob . . .

gregmchugh

Joined: 03 Apr 2012
Posts: 42

Posted: Tue Apr 10, 2012 7:37 am Post subject: Re: Open source product development for OBAM aircraft

Bob,



Looks good. Let's try to keep the option open to drop in the PICAXE 8 pin

chip in place of the standard PIC chip. I am not sure of the PIC chip they

started with but the pinouts for the current PICAXE 8 pin are shown here.



http://www.picaxe.com/docs/picaxem2.pdf



I believe the PICAXE is simply one of the standard PIC chips with

their firmware installed and the chip marked with their designation,

at least that is what they used to be. They used to still have the

PIC product designation on the chip with a sticker applied to

show the PICAXE designation.



The standard PICAXE development boards connect pins 2 and 7 to

the serial connection to the PC and the onboard loader is connected

to these pins. In practice the pins can be reconfigured in the

application so they are available when off of the development board

for other uses. If they are not used in the application then the option

to keep the serial port connected to the PC would still be available

even in the application hardware configuration. Connecting these

to the external connector would allow access to the chip from a PC.

Of course the serial port could also be used for I/O to any

standard RS-232 serial device by the application. Here is a description

of the 8 bit development board.



http://www.picaxe.com/docs/axe021.pdf

Greg McHugh

nuckolls.bob(at)aeroelect
Guest

Posted: Tue Apr 10, 2012 2:57 pm Post subject: Open source product development for OBAM aircraft

At 10:37 AM 4/10/2012, you wrote:

 	  Quote: 	
		  



Bob,



Looks good. Let's try to keep the option open to drop in the PICAXE 8 pin

chip in place of the standard PIC chip. I am not sure of the PIC chip they

started with but the pinouts for the current PICAXE 8 pin are shown here.


	


  Obviously, we have options to do anything. I'd like to see what

  we can do with the PIC12F683. It's a part I stock in quantity

  and it fits the present board layout. It's certainly capable

  of the sorts of jelly-bean products that would fit into the

  proposed enclosure.



  I am considering a next-step-up which will have to have USB

  interface capability. The PIC182458 offers 12-bit a/d and

  USB2.0 support on chip but that's a do-lots-board still looking

  for some projects. Right now, the 12F683 board is done per the

  schematic I posted and easily modified to handle the larger

  power mos-fets for the 5A wig-wag controller.



  Lets get our feet wet in a short path project first . . .

  are there pin-for-pin subs for the 12F683 in PACAXE?



   Bob . . .

nuckolls.bob(at)aeroelect
Guest

Posted: Tue Apr 10, 2012 3:08 pm Post subject: Open source product development for OBAM aircraft

At 10:37 AM 4/10/2012, you wrote:

 	  Quote: 	
		  



Bob,



Looks good. Let's try to keep the option open to drop in the PICAXE 8 pin

chip in place of the standard PIC chip. I am not sure of the PIC chip they

started with but the pinouts for the current PICAXE 8 pin are shown here.


	


   I looked up the pinout on the PICAXE-08M2. Seems they swapped

   sides for the a/d input pins . . . we can do a board that will

   exploit the 08M2 but it would be a different board.



   Bob . . .

nuckolls.bob(at)aeroelect
Guest

Posted: Tue Apr 10, 2012 3:38 pm Post subject: Open source product development for OBAM aircraft

At 10:37 AM 4/10/2012, you wrote:

 	  Quote: 	
		  



Bob,



Looks good. Let's try to keep the option open to drop in the PICAXE 8 pin

chip in place of the standard PIC chip. I am not sure of the PIC chip they

started with but the pinouts for the current PICAXE 8 pin are shown here.


	


   I looked up the pinout on the PICAXE-08M2. Seems they swapped

   sides for the a/d input pins . . . we can do a board that will

   exploit the 08M2 but it would be a different board.

   Scratch the above. Got my tongue tangled around my eyeteeth

   and couldn't see what I was typing. Seems the 08M2 is on

   a PIC12F1840 which is pin compatible with the 12F683. So

   we could sure go with that one if suits your programming

   preferences.



   Bob . . .

Eric M. Jones

Joined: 10 Jan 2006
Posts: 565
Location: Massachusetts

Posted: Tue Apr 10, 2012 3:40 pm Post subject: Re: Open source product development for OBAM aircraft

 	  Quote: 	
		  >One question I forgot to ask in the previous note. When I looked at

>using the Perihelion Wig-Wag for my LED landing lights I was

>surprised at the compact size and the capability to switch the loads

>on standard landing lights. Is there something unique about this

>design to handle the 250 watt load?



   You'd have to check with Eric as to his

   design philosophy but based on what is

   generally known, getting the heat in the

   FETs under control has two components…. 	




Bob is right. The Perihelion Design Wig-wags use either an FDD6670S (30V) or an FDD8444 (40V). Both of these are “D” packages (TO-252) which are surface mountable and about 1/4” square. The mounting is done by soldering the package to the PCB which itself is designed for maximum dissipation. The Faston lugs themselves are included in the thermal calculations, as is the potting compound.



For this design I chose “high-side” switching, since the loads are remote and might be remotely grounded. The difference is that “low-side” switching would normally require an additional wire to the load. The wig-wag itself weighs less than the additional wire would. The disadvantage is that more a complicated driver for the FET gate is required. 



Mosfets are wonderful parts for wig-wagging, where there is no need for an intermediate state, they just switch ON and OFF. But they have to be operated within their specified range. This was a problem when people applied these wigwags to HIDs. Most HIDs have an extremely high ignition voltage, then a high boost voltage to induce IR heating to the anode terminal of the lamp. The boost current is over 30A for a substantial fraction of a second. This would wipe out the FET. The solution was to supply a CL-11 surge protector and a Zener across the line for those who intended to wigwag the HIDs. Now they wigwag well.



But for incandescents (or LEDs), and using a surge protector, the wigwag will supply 20A per pulse. This is possible because the gate driver fully saturates the gate allowing the lowest possible Rds(on) resistance, and thus the smallest I x I x Rds(on) heating. For 20A, the heat dissipation is about 3.2W, split among two devices!



Early versions used a 555 timer and a mosfet driver. Later versions have used a small RC oscillator and a CD4060 CMOS 14 Stage Ripple-Carry Binary Counter/Divider. This would seem an unusual part for this application, but it simply divides the RC oscillation and goes tick-tock at the correct duty cycle and timing—it does all the work. The 555 was hard to make it do 50% (but it could be done with a few more parts) and the first pulse was always 1.5X longer than any succeeding pulse…which is inelegant and stressed the fets. I was never satisfied with the early design on this matter.



All parts are surface mount and are assembled under a stereo microscope. The entire PCB is 1.16” x 0.90”. Early on I got a note from a customer saying, “Okay, I see the switch, but where is the wigwag circuitry?”  There are several styles of this wigwag, one with the switch WW(d) and one BYO-Switch WW(c) that allows all sort of switching configurations. I also make these in 12 and 28V and a special double rate for single head-lights (e.g. the Europa or Glastar). And I supply the add-on bits for HIDs.



But Bob's approach for flashing LEDs is good.

ww(d).jpg

Description:

Filesize:

525.16 KB

Viewed:

14273 Time(s)

_________________
Eric M. Jones
www.PerihelionDesign.com
113 Brentwood Drive
Southbridge, MA 01550
(508) 764-2072
emjones(at)charter.net

nuckolls.bob(at)aeroelect
Guest

Posted: Tue Apr 10, 2012 7:39 pm Post subject: Open source product development for OBAM aircraft

At 10:37 AM 4/10/2012, you wrote:

 	  Quote: 	
		  



Bob,



Looks good. Let's try to keep the option open to drop in the PICAXE 8 pin

chip in place of the standard PIC chip.


	


   I've been poking around the 'net trying to get a better

   picture of what the PICAXE is all about. If I am reading

   the postings right, PICAXE chips have an on-board interpreter

   for tokenized BASIC commands . . . these tend to be very

   slow with respect to what the chip can do with optimized

   object code.



   I saw some links for a basic compiler that targets

   PIC devices.



http://www.sfcompiler.co.uk/swordfish/



   I think this does a righteous compilation of

   the BASIC source code for much more efficient

   use of memory and greater speeds. I'm going

   out of town tomorrow with Dr. Dee. She has

   a conference to attend in Oklahoma City and

   I'll be driving her and three of her students

   to the conference. I'll have quite a bit of

   free time while they take care of business.

   I'll look a bit deeper into the Swordfish

   product and . . . there are undoubtedly

   others.



   I can see the allure of the PICAXE product

   for neophytes to get their juices flowing

   for herding bytes around in silicon. It would

   no doubt suffice for the wig-wag project

   too.  But a serious compiler would let you

   run less expensive chips at their best

   speeds. I think there'a a place for both

   sets of tools.



   Since both tools target PIC products, the

   board layouts for things like the do-much

   architecture will work across both tools.

   Bob . . .

JLuckey(at)pacbell.net
Guest

Posted: Tue Apr 10, 2012 10:02 pm Post subject: Open source product development for OBAM aircraft

Bob,



Your assessment of the Picaxe is right on.  They are great for

beginners/simple/low performance devices but their interpreter is at least a

couple of orders of magnitude slower than devices programmed w/ a compiler.



There is a myriad of experimental aviation applications where that level of

performance would be adequate.



Several years ago when I first started programming uControllers, I examined

PICAXE & Parallax Stamps.  I settled on the Stamps (They also have built-in

interpreters).  They are great for getting up and running simply & quickly,

which is key for the initial learning curve, but I quickly outgrew them.



As I'm sure you are aware the standard in BASIC programming w/ PICs is

microEngineering Labs PICBASIC PRO Compiler.  It is a great product and

reasonably priced at $120-$270.  I know several people who use & really like

it.



And, of course, Real Programmers program uCons in C 



-Jeff



--

pmather

Joined: 17 Sep 2009
Posts: 33

Posted: Tue Apr 10, 2012 11:07 pm Post subject: Open source product development for OBAM aircraft

Bob



Picaxe is slow compare to compiled code - of course. However, this is more

than made up for by ease of coding in many circumstances. Facilities like

ADC, I2C,Serial,PWM etc are all implemented as single basic commands.

Running a 08M2 at 32Mhz gives plenty of speed for most applications.



I also use Swordfish when speed is an issue but this only works for 18F

series chips (smallest is 18pin). Sourceboost works for 12F and 16F as well.



In both cases hardware handling is at a much lower level so you need to

spend much much more time with the PIC datasheet.



Picaxe has a great support forum www.picaxeforum.co.uk



Connection of the picaxe to a PC with a serial port for programming requires

two resistors - nothing else, so with an 08M2 you can be up and running with

a bit of stripboard for less than $2.



Try it - you'll like it!



Best Regards



Peter



--

gregmchugh

Joined: 03 Apr 2012
Posts: 42

Posted: Wed Apr 11, 2012 7:07 am Post subject: Re: Open source product development for OBAM aircraft

Bob and all,



I agree with the plan to start with the PIC12F683 since you have those

in stock and the board is designed. As you noted there is a drop in

PICAXE replacement for those interested in that option. I am

planning to code the wig-wag function using the standard Microchip

development tools using C. I ordered the PICkit 1 Flash Starter Kit

directly from Microchip on Monday for $36 plus shipping and it is

scheduled to arrive today. See Part Number: DV164101 at

http://www.microchip.com/. There are some newer development tools

that can handle newer PIC devices but this is a good kit to get 

started with the PIC12F series



I checked on the SWORDFISH Basic compiler and it appears to only

support the 18F family, 12F does not seem to be supported.



As noted by many there are lots of options for software development

from assemply code (which is what real men use, higher order

languages are for wimps), compiled C and BASIC, interpreted BASIC,

and code from flowcharts. Lots of options are good, an advantage to

using PIC, and I am sure there are more that have not been mentioned.

Any single software development solution would not target a wide

enough user base to make these types of modules viable for anything

but limited use.



I am very pleased to see that there are more responses coming in

from software folks, I was afraid I was stuck in a sea of hardware folks.

I won't start with the war stories on embedded software development.

Here is a Dilbert that shows how the war story saga ends...



Anyway, welcome software folks and let's here lot's of ideas on

what you would like to see happen with this project...



Greg McHugh

28703.strip.gif

Description:

Filesize:

37.06 KB

Viewed:

14247 Time(s)

nuckolls.bob(at)aeroelect
Guest

Posted: Thu Apr 12, 2012 2:44 pm Post subject: Open source product development for OBAM aircraft

At 10:07 AM 4/11/2012, you wrote:




Bob and all,

I am planning to code the wig-wag function using the standard Microchip

development tools using C. I ordered the PICkit 1 Flash Starter Kit

directly from Microchip on Monday for $36 plus shipping and it is

scheduled to arrive today. See Part Number: DV164101 at

http://www.microchip.com/.

   Okay. I I'll get the FAT fets laid onto the artwork, modify

   the do-lots schematic to agree, and get some boards

   ordered.

I checked on the SWORDFISH Basic compiler and it appears to only

support the 18F family, 12F does not seem to be supported.

   Hmmm . . . fooey. Keep us up to date on future recommendations/

   discoveries. I used to write a lot of 6800/6502 assembler for

   embedded applications. I use Turbo Basic for test systems. I'd

   like to get salty with the RSC controllers but after one had

   the 150+ constellation of instructions in the legacy chips,

   it takes a whole new outlook on programming to get 'er done

   with less than 40!



Any single software development solution would not target a wide

enough user base to make these types of modules viable for anything

but limited use.

   agreed



I am very pleased to see that there are more responses coming in

from software folks, I was afraid I was stuck in a sea of hardware folks.

I won't start with the war stories on embedded software development.

Here is a Dilbert that shows how the war story saga ends...



Anyway, welcome software folks and let's here lot's of ideas on

what you would like to see happen with this project...



   Had to put new tires on the car today so didn't get to

   'play' . . . I'll have some time in the morning.

   Bob . . .

gregmchugh

Joined: 03 Apr 2012
Posts: 42

Posted: Thu Apr 12, 2012 3:31 pm Post subject: Re: Open source product development for OBAM aircraft

Bob,



As noted by Peter in an a previous note Sourceboost does provide

C and BASIC compilers for PIC. I really have no prior experience

with the PIC so I don't know the strengths and weaknesses of

the various options. I just went with the Microchip solution since

it gets you going for the least amount of money and is supported

by Microchip. The free version of the compiler that comes with

the Microchip IDE does not include all of the code optimizations,

you have to pay to get the full featured version.



Greg McHugh

gregmchugh

Joined: 03 Apr 2012
Posts: 42

Posted: Fri Apr 13, 2012 5:17 am Post subject: Re: Open source product development for OBAM aircraft

Bob,



Could you post a higher resolution version of the schematic that

you provided in a previous post. It is hard for me to make out

the details.



The PICkit 1 Flash Starter Kit arrived from Microchip and I cleared

some space on the bench, set up a PC to use for software development,

connected the starter kit via USB (the LED's flashed as they should),

loaded the Starter Kit software from the CD provided (this does the

downloads to the PIC12F675 provided on the kit board).

I also downloaded the latest Microchip MPLAB IDE (version 8.84)

from the web (the CD provided is an older version). Checked that

the download worked by loading the first tutorial lesson code and

that seemed to work just fine. All-in-all no real problems getting

things up and running. I will write up a short tutorial on how to

get started for anyone who wants to do it.



I will take a look at the tutorial lessons in the next few days. In

fact the first tutorial covers switch debouncing and controlling

LED's so I expect we have the core for the wig-wag software right

there.



Greg McHugh

Display posts from previous:

	Matronics Email Lists Forum Index -> AeroElectric-List	All times are GMT - 8 Hours
Page 1 of 1

Jump to:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum