triise's A-10C pit

[triise]

Letters are better . and while in general that front panel is usable the fat line around the 3 indicators is quite distracting . The reason it is drawn as it is in my dxf is due to how areas to be engraved are defined for laser. It seems that in your case you would need to modify it to be single line. Same applies in lesser extent to the box around "tems ". Let me know if you need help with that.

 

Yes I know. The fat line is a result of me forgetting to delete just that machining line in the g-code. It's been machined with a 2mm end-mill, not the v-shaped engraving tool it was supposed to. I't was late in the night and this was the last of the engraving panels, so I couldn't make a new one. Have new engraving plates on order and will make a new one when I get it. :)

 

Also the text and boxes are a bit offset due to the fact that I moved the machining origin and forgot to note the coordinates. The result was that I had to try to align in manually / visually...

 

Hopefully the shipment from cnc-plus.de will be here when I'm back home (They are not the fastest supplier), so I can make a new one.

 

Thanks for the feedback :thumbup:

[triise]

I had some light bleed through issues around the edges of mine even after painting. I ended up using liquid electrical tape. It is much thicker than paint and doesn't bleed between the layers. It's also a brush on paint so you don't have to tape off the panels to prevent overspray.

 

BTW I have Reactor1's panels too. He really did a nice job on them.

John

 

I've tested the light bleed with this paint and it looks like it will be ok. It's a kind of plastic paint which isolates the light pretty good.

 

T

[triise]

Ok, so i've finished the landing gear panel, started the UFC, Caution Light Panel and the CMSP. Planning to mill a custom PCB for the Caution light panel, will be the task at hand tonight. Also, the UFC doesn't look pretty on the inside, but that's not the important part

I just glued the tactile switches on the back of the acrylic.

Attached is some of the pictures. Will post more detailed later, as I'm on my phone right now.

 

T

[agrasyuk]

Aside of couple neat-picks this is looking good.

Don't skip on backlighting !

Warning panel will not work as is. you need to figure out some way to insulate the light bleed between individual anunicators.

I'm not sure if one arduino board will be able to power the entire warning panel

[Warhog]

This is what Anton is talking about.

 

A framework to divide the area into 48 cells.

 

 

Then you get this. Only one cell being illuminated

 

 

 

 

One Arduino board can handle the entire Caution Panel assuming that you lay out the circuit board as a matrix. You can use a MAX7219 chip to direct the output from the Arduino board to all or one or any combination of LEDs and you won't need a resistor for each LED. If you attempt to light every LED from an Arduino board "as is" the current draw will fry the board. 48 LEDs x 20mA = total current of 960mA

[triise]

@Warhog: a matrix won't draw any less power than my setup. 20 mA is 20 mA any way you connect it. I'm using 1k ohm resistors on every led. Calculated current draw is 7 MA. Total of 336 mA when all the LEDs are being lit. That should be within the Arduino limits according to their page.

 

Anyway, thanks for the tip on light bleed. As you can see, work is in progress.

 

As for the backlighting, I'm using led strips turned upside down and glued to the back of my panels (which is all white acrylic). Will post pictures later.

[triise]

New casing for the LEDs on my caution panel. Looks better.

Soldering the PCB

[ClayM]

triise

 

your correct 20 mA is 20 mA. But what if you only turned on one LED at a time but did it very quickly. What would your total current draw for the arduino be?

 

Listen to Warhog. He would have saved you the hassle of all those resistors!

 

I tried your way without much success.

 

Warhog, maybe you could share a sketch that would work with DCS-BIOS and MAX7219 to clarify it for everybody!:smilewink:

[cefs]

Hi triise, not sure what figure you are quoting when you say 336ma is within spec. The max current draw is around 200ma for the chip, you may be looking at 500ma that can be supplied by usb or the the current that can be supplied by the voltage regulator, but that is current that can be consumed by the entire board. Chips like the MAX 7219 use pwm to reduce current draw, there are plenty of different chips out there, but definately the way to go.

[triise]

Hi triise, not sure what figure you are quoting when you say 336ma is within spec. The max current draw is around 200ma for the chip, you may be looking at 500ma that can be supplied by usb or the the current that can be supplied by the voltage regulator, but that is current that can be consumed by the entire board. Chips like the MAX 7219 use pwm to reduce current draw, there are plenty of different chips out there, but definately the way to go.

 

http://playground.arduino.cc/Main/ArduinoPinCurrentLimitations

 

Here is some info from the Aruduino site.

[cefs]

Hi triise, you need to look at the datasheet supplied by Atmel, I assume its a 328p your using. The datasheet says 200ma. If you wanted to keep your existing setup you could use some uln2803's.

[FSFIan]

Hi triise, you need to look at the datasheet supplied by Atmel, I assume its a 328p your using. The datasheet says 200ma. If you wanted to keep your existing setup you could use some uln2803's.

 

According to the Arduino site it is 200 mA per ground pin. Since the DIP version of the ATMega328p has only one VCC and one GND pin, that is also 200 mA for the whole chip if your Arduino board uses a DIP package.

 

The surface-mount packages have more VCC and GND pins, so if I interpret that correctly they can sink and source a total of 400 mA as long as it is properly distributed over the I/O pins. The Mega can apparently deal with up to 800 mA.

 

However, these are absolute maximum ratings, i.e. Atmel guarantees that as long as you stay below that, your chip won't fry. For reliable operation, you may have to take other factors (such as heat dissipation) into account, I have no idea if that can be a problem here.

[cefs]

I think the Arduino site has it wrong. The Atmel manual states 200ma between VCC and GND. As power and ground rails will be tied internally then that will be total current consumption regardless of the number of VCC and GND pins. It is never a good idea to drive loads directly, even leds. One or two are fine, and that is why general purpose micros can drive them, but they were never designed for that purpose. It is always best to switch with transistors/fets instead. Pushing 800ma through a micro to me sounds crazy, think they really need to have a look at that.

[triise]

I'm planning on using a Mega to control this panel. And as far as I can see it can supply 800 mA. Then I should be well within the limits. Anyway, I ordered some matrix chips from eBay, so if it doesn't work I'll switch to that solution.

[cefs]

My advice would be to not do that, but if you do then use extreme caution and think about personal safety above all else.

[Boltz]

If current with arduino is a problem then I would just put in a few Darlington transistor arrays. Don't require any matrix code. Just drop them in line with arduino connections.

[triise]

Thanks, I´ve just ordered some Darlington arrays now. Will have to wait for them to arrive from the UK before I start again.

 

Next up is the CMSP panel. I´ve read a bit about 16x2 vs 20x2 LCDs. Currently I have 2 16x2 panels, but I understand that it will be best with a 20x2 and I have some on order from eBay. Don´t want to start cutting before I have the panels (as the dimmensions probably will be a bit different).

 

I´m also looking into the Engine gauges, but can´t seem to find how to address the servo motors on a PCA9685 PWM driver. Will have to do a bit more google... :pilotfly:

Edited March 18, 2015 by triise
typo

[Warhog]

 

Listen to Warhog.

!:smilewink:

 

That may not be the best advice as I'm often as confused as most everyone here looking for answers.:megalol:

 

 

Warhog, maybe you could share a sketch that would work with DCS-BIOS and MAX7219 to clarify it for everybody!:smilewink:

 

My caution panel is still under construction so i cannot attest to how well the MAX7219 works with an LED matrix. I have used it on numerous occasions to drive 7 seg displays for my Tacan, ILS and UHF Repeater. I will definately be using it to drive the Caution panel.

 

I would first get the data sheet for the 7219 and read it carefully. It's easy to install but there is important info on decoupling capacitors and setting max current. Then see the Arduino site as they have libraries for it. Lastly, refer to the last page of the DCS-BIOS Users Guide as there is the beginnings of a sketch to run the matrix.

 

Then after you've done all that, let me know if it works and I'll use it as well.:music_whistling: :lol:

 

Kidding aside, my matrix is already in place and my design of it will not work with the example Ian provided in the Users Guide. My sketch had to remap all of the indicators because I screwed up in laying out the PCB. :doh: So it won't work with your matrix once you've built it. I do believe Ian's sketch is the solution your looking for.

[cefs]

The Max7219 is great for driving seven seg leds, but my concern would be whether it can produce enough brightness for each of the caution indicators. At least with darlington arrays you have much more control over voltage/current and could easily have 2 leds per channel. The MAX7219 though I beleive has adjustable brightness built in which could be useful as long as it can be bright enough at max setting, also the number of pins needed to control it a lot less.

[FSFIan]

In addition to a steady-state current rating (typically 20 mA), most LEDs have a higher peak rating (e.g. 40 mA) for a given duty cycle. In other words, if you are using LEDs in a matrix, you are allowed to operate them with higher currents because they have time to cool off again in the time their row is not being driven. This can compensate for (some? all? of) the loss of brightness caused by being on only a fraction of the time.

 

Check your LED's datasheet.

 

If you are worried about brightness, do your caution lights need to be brighter than your 7-segment displays? And if you do need more brightness, wouldn't it be easier / more cost-effective to go with a different type of LED instead of a more complex circuit and PCB layout?

[cefs]

Ian, very good questions, and I do not know the answers, my concern really is down to the area that a single led has to back light, will a single led of any variety driven by a MAX7219 be able to light that area sufficiently. I would want each indicator to have a uniform and bright display, it needs to be bright enough for you to notice and not just a 'hot spot' in one area. My initial thoughts are that 2 leds would be better.

[Boltz]

Compensating for brightness by using a peak current will not make them brighter. Since they are flashing your eyes will merge it together and they will not be as bright. That is basically a pwm circuit which is used for dimming LEDs. Flashing them will decrease brightness and correcting that with higher current will make them back to about the same level. If you do not need the matrix to decrease the controller pins then darling tons are the easiest and most customisable way to go. The current is more easily controlled, easier to debug and they are simple to implement.

[Warhog]

Ian, very good questions, and I do not know the answers, my concern really is down to the area that a single led has to back light, will a single led of any variety driven by a MAX7219 be able to light that area sufficiently. I would want each indicator to have a uniform and bright display, it needs to be bright enough for you to notice and not just a 'hot spot' in one area. My initial thoughts are that 2 leds would be better.

 

The creation of hot spots is dependent on how you design the caution panels "cell structure" and the material used for the face plate and not on how many LEDS you use. Even my first caution panel used only one LED per cell.

 

This is a picture of my first Caution Panel. In the dark it was actually too bright with all the LEDs lit, but there were no hotspots. It used a lattice structure to create the cells as does my second panel. The lattice is made from white styrene. To create a type of diffusion filter, to ensure there were no hot spots, I sanded the back of the acrylic faceplate with a fine paper (600grit) and that created the diffusion. It worked well but it could have been better.

 

 

This is my second (and latest) caution panel build. It uses only one LED per cell. It also has a very thin sheet of white plastic that acts as a diffusion filter. When all assemble the brightness is perfect.

 

 

It is quite bright when in a well lit room and we are only talking 20mA. But it needed to be bright so I could added a diffusion sheet. It's now perfect in both normal room light and a dark room. As with my first panel the underlying structure is built as a lattice of white styrene plastic which does a marvelous job at reflecting the light throughout the cell which also contributes to diffusion.

 

 

And finally, you can see the diffusion layer and the effect it produces. No hot spots. I use a diffusion sheet between the LED and the Acrylic faceplate as below. This diffusion sheet is just a sheet of .02" thick white styrene. It comes in various thickness so its easy to try different thicknesses to find what works for you. This only works as I described by using clear acrylic painted black and engraving through the paint. You could use engraving plastic but then the LEDS would need to be brighter and the diffusion sheet deleted from the build.

 

 

 

And here it is almost complete...need to painted the edges black, drill a few holes and clean it up a wee bit.

 

 

It's always good to see how others have built their panels. I find it helps generate new ideas and hopefully more creative approaches to solving these kind of build problems.

[cefs]

Some mighty impressive work Warhog, and good to know the MAX chip is up to the job. I did come across this yesterday https://www.coolcomponents.co.uk/white-led-backlight-module-small-12mm-x-40mm.html so it is clearly evident that a single led would suffice.

[triise]

20x02 CMSP LCD up and running!

Wrong colour and all, but got them cheap second hand. Will probably get some green ones later.

 

Now I'm waiting for the Darlington arrays to arrive.