help on DC-AC inverter for Public art Project. Thks

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Fri Aug 22, 2008 7:59 pm

Hi Guys,
got it, I'll put the packages on the mail first thing on Monday.

Rob, Thanks for the new diagram!, I normally use Jameco, Digikey, all electronics (for sales), Allied, the usual.

Dyaker, Thanks a lot for the Database and the corrections!
Thanks Guys!!

MrAl, You can find info of the project at www.federicomuelas.com/blue_flower

Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Fri Aug 22, 2008 10:28 pm

Hello Federico,

Thanks for the list of suppliers. I will use parts from these places when picking components for the PCB outlines. I should have a gerber file layout of a first draft of the receiver-PIC board by early next week. Will probably not be the final layout as there are probably still some changes to come, may use a different PIC to the one shown.

Regards
Rob
LIFE….the crappy bit between birth and death

dyarker
Posts: 1709
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Sat Aug 23, 2008 2:04 am

Viking,

Just noticed on timing diagram you put 2 stop bits for serial data characters.

I will hard code the number of stop bits instead of putting option in a .ini file because user changing it would upset overall timing. Doesn't matter to me either way, just don't want to change it later. If you need 2 stop bits for PIC code, okay. If you can get by reliably with 1 stop bit, the frame rate can be about 9% faster.

The number of data bytes is now 472 (In data I posted on my site, numbered 0 to 741). Do you need padding to 512?
---------------------------
Each frame will have it's own display time ranging from about 175mSec (time to send data) upto about 200 Seconds. There will be NO CHANGE in data sent; the program will just wait longer after last data byte before sending break for next frame. Okay?
---------------------------

Federico,

What kind of remote monitoring via network do you have in mind? It won't be implemented in version 1, but I need to leave "hooks" in the program so it is easier to add later.

C U L -
Dale Y

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Sat Aug 23, 2008 9:30 am

Hello Darker,

Yes, there is one start bit followed by eighth data bits followed by two stop bits, as in the DMX specification. At the moment I can't think of a reason for using this arrangement, but it is quite common to have this arrangement, so there must be a good reason for it.

The actual number of channels should not matter. Once you are into the 'mark-between-packets', all PIC's know that there are no more data channels in the current data packet and wait for the 'break' to start counting channels again. The 'mark-between-packets' could conceivably be any length, as the PIC's will simply wait for the 'break'.

Do you mean 200mS instead of 200 seconds?

By the way, in a Cat5 cable with 4 pairs, are the pairs in pin order, e.g. 1&2 twisted together, 3&4, etc.
Regards
Rob
LIFE….the crappy bit between birth and death

dyarker
Posts: 1709
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Sat Aug 23, 2008 11:50 am

Okay, 2 stop bits it is.

Small update at http://www.yarker-dsyc.info/BlueFlower/

Re: your Cat-5 question, here are two of my old pages that may help:
http://www.yarker-dsyc.info/Notebook/Te ... abl01.html
and
http://www.yarker-dsyc.info/Notebook/Te ... abl02.html
(don't confuse my end "A" and end "B" with T-568-A and T-568-B specifications.) www.wikipedia.org also has a lot of good technical pages.
Do you mean 200mS instead of 200 seconds?
Nope. He'll (or UNM user) be able to set frame display time from about 174mSec (break plus data time) to over 200 Seconds (the number of 100s of nanoSeconds that will fit in a signed 32 bit variable. Each frame can have a different display time. I'll put these specs with "why" at http://www.yarker-dsyc.info/BlueFlower/ At nine pages it is getting hard to find stuff here! In fact do I have your permission to copy/modify and post a couple of your diagrams on my site?

C U L -
Dale Y

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Sat Aug 23, 2008 12:22 pm

Hello darker,

OK on the 200seconds, the PIC will wait for the next 'break' however long that is.

On the CAT-5 interconnection, I've made a mistake on the schematic, which I will need to correct. On the main board (receiver-PIC), there is an 8 way output to the next board in the chain via a simple line driver. The next display module will have an 8 way input connector to a simple line receiver and a 8 way output connector, again via a simple line driver. So each display module acts as a simple 'repeater'.

But I'm unsure of the connection configuration I should use. There is an 'A' and 'B' version, but in terms on physical connection, they are identical, it is just the colours that change. Is this correct? I want to see if I can devise a connection method that will be fool-proof, that is no damage will occur if the cables between modules are swapped so that two outputs are connected together.

I guess I should include terminators at the input of each module, as each is in effect 'at the end of the line' as far as its signal is concerned.

Thinking about this, it would probably be possible to drive the whole display directly from ONE receiver-PIC board if the PIC on that board were set to clock all data from channel 1!

Regards
Robert
LIFE….the crappy bit between birth and death

dyarker
Posts: 1709
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Sat Aug 23, 2008 9:02 pm

Right, between 568-A and 568-B, only the colors change. Pins 1 and 2 are a pair, pins 3 and 6 are a pair, pins 5 and 4 are a pair, and pins 7 and 8 are a pair. The white wire with color spots is always "tip", the solid color wire is always "ring". For RS-485 "tip" would be 0V for mark, 5V for space; "ring" would be 5V for mark, 0V for space.

One PIC per panel? Or one PIC per A6818? I thought one PIC per panel, driving however many A6818s as needed for a panel. (Like block diagram I put on page 7 except I put strobe from wrong source.) Are clock and strobe out of PIC converted to RS-485 then back to 0-5V logic level for each A6818? If logic level for clock and strobe to A6818s then "end-of-line" is ONLY on last A6818.

If a PIC for each A6818, then only last PIC (with RS-485 receiver) is "end-of-line". (implied by your diagram on page 7)

I think I need to change my block diagram till you say "Yes, that's it!" Hints please.
Dale Y

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Sat Aug 23, 2008 10:27 pm

dyarker wrote:One PIC per panel? Or one PIC per A6818? I thought one PIC per panel, driving however many A6818s as needed for a panel. (Like block diagram I put on page 7 except I put strobe from wrong source.) Are clock and strobe out of PIC converted to RS-485 then back to 0-5V logic level for each A6818? If logic level for clock and strobe to A6818s then "end-of-line" is ONLY on last A6818.

If a PIC for each A6818, then only last PIC (with RS-485 receiver) is "end-of-line". (implied by your diagram on page 7)

I think I need to change my block diagram till you say "Yes, that's it!" Hints please.
Hello Dyarker,
sorry to add confusion, this was just speculating idly, nothing has changed, the scheme is still the same with one PIC per panel and a single RS485 line connecting all of them. The 'line drivers' are still operating 0-5V, I just did not like the idea of what would effectively be the 6818 driving the cable to take the Data Out signal to the next stage. But the scheme shown in your drawing would work just as well with the driver on the receiver-PIC board driving a common line to all the other boards in the chain.
Only difficulty would be that each 6818 would drive the line to the next board in the chain (the data out line would not be 'common' to all the boards). So I think I will stay with my intended scheme.
I will post a block of the system to show what I mean.

Regards
Rob
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Sun Aug 24, 2008 6:25 am

Hi Guys,

I don’t know what happen with the last post I put yesterday, It didn’t come through I guess.
Dyarker, thanks a lot! I’ll post tomorrow a list of task that the computer need to perform so you can get the bigger picture in case you need to add more hooks to the code (i.e. find the shadow of the building cast over the screen to define the line between the projected and the rendered image) everything will be clearer when you guys watch the presentation DVD (a lot of renders and stuff).

Rob, Thanks a lot! I’ll try the layout with the milling (and with the laser too) as soon as I receive your Gerber file, even if it’s not the final one, to check if this system will work.

By the way, next week, the New Mexico Arts Commission and the University are hanging a large scale render of the project on the Concourse A at Denver Airport (during the DNC) to promote the piece, as one of the five NM public art projects selected, let’s see if we get some feedback.

Thanks guys!

Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Mon Aug 25, 2008 5:43 am

Hello Federico,
I'm ready to send you the preliminary gerber file for the Display-receiver-PIC board. Can you reply to e-mail so that I can forward you the attachment.
Regards
Rob
(added:- here is a jpg layout, showing pads tracks (red) and links (white). Note the slight discontinuity in the numbering of the pixel outputs, this layout results in a much easier routing layout. Note that the files I sent you are for the solder side of the board. As all gerber plots are normally viewed from the component side, the tracks must be 'plotted' in reverse (mirrored). If your CNC program cannot reverse them, then I can plot them reversed).

Image
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Mon Aug 25, 2008 10:14 am

Thanks Rob!!!
Got everything, I'll give it a try this week and I'll let you know how it goes.
Thanks a lot
Fede

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Aug 26, 2008 1:52 pm

Hi guys,
these is the task list, we have been working on the first (and most important) one, but hopefully the rest will be eventually implemented on the piece. It might be a little bit confusing but everything will be clearer when you get the DVD (I think). Also I add an extended description of the project at the end of this post (I don’t know if I did this already).
thank you so much Guys.

----------------
TECHNICAL GOALS BY PRIORITY ORDER:

Goal 1- ink tracking and convert video/image to serial protocol. Analyze the moving image of the ink expanding into clear water; derive the pixilation information and design and implement serial control protocol. The image will be feed into the system by a camera aiming at the water container inside the ink projector. The screen is made of 3,700 pixels distributed in a circular grid of 70 x 70 units; therefore the final resolution of the video to analyze will be 70 by 70. The rate of the speed will be 4 frames per second max.

Goal 2- shadow tracking. Analyze the image of the screen and determine where the building shadow is being cast and then determine an accurate segmented line formula representing the shadow, this will be use to set the mode of the pixels on the screen, reflective or rendering mode. If feasible other more complex shapes will be tracked as well such us the shadow of a cloud or a tree.

Goal 3-Provide a means to visualize, debug and maintain the system from a remote location via an Internet connection. Provide an interface to the university so that custom messages or low-resolution bitmap movies can be displayed on the screen. This interface must be fool proof and easy to use. If possible, the real time video of the screen and the ink will be broadcast on the project web.

Goal 4-Provide a means to translate the image of the drop of ink into sound, so the vertical position of the shape modifies the pitch of the sound and the horizontal the panoramic value. The sound will be distributed to several sound connectors installed on the ink machine or the amphitheater.

Goal 5-Provide an e-mail based control system to warn about possible occurrences on the piece (ink level, water leaking, excessive temperature, etc).

Goal 6- Provide an interface to change the different modes of the piece: normal mode (art piece), board mode (external source digital moving or still images rendered by the screen), and digital projection mode (video projector inside the ink machine bypassed).
----------------
Description:

Blue Flower/Flor Azul:

Public Art Piece for the George Pearl Hall Building, School of Architecture and Planning at the University of New Mexico, Albuquerque (designed by Antoine Predock)

“Blue Flower/Flor Azulâ€

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Aug 26, 2008 1:56 pm

Rob,
I downloaded Ranger 2 as you recommend me, is a great program!!, thanks!
(learning it right now)Could you please send me the Part List that you use to create the layout? Or some kind of reference so I can order the parts online.
Thanks a lot!
Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Tue Aug 26, 2008 2:42 pm

Hello Federico,

I will send you the design files to your email. I create most of the component outlines by hand with every design, the component library is not particularly useful and the symbol library is a bit on the 'chunky' side as well so will always use modified versions of the standard library. So I will also send you a list of the components that I based the outlines on (that's the reason for asking about component supplies that you generally use). Remember that this is a test layout. I have since rearranged the layout to make the pixel output all contiguous. I just want to find out what is the resolution of your CNC machine really is.

Regards
Rob
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Aug 26, 2008 5:49 pm

Hi Rob,
Thanks!
Let's do this layout then with the CNC to see what we can get.
Fede

Post Reply

Who is online

Users browsing this forum: No registered users and 10 guests