All About Circuits Forum  

Go Back   All About Circuits Forum > Electronics Forums > General Electronics Chat

Notices

General Electronics Chat Discussion forum for general chat about anything electronics related, including asking questions about material in the All About Circuits E-book, Worksheets, and Videos.

Reply   Post New Thread
 
Thread Tools Display Modes
  #51  
Old 04-28-2012, 12:50 PM
Bill_Marsden's Avatar
Bill_Marsden Bill_Marsden is offline
Super Moderator
 
Join Date: Mar 2008
Location: Dallas, TX (GMT-5 w/ DST)
Posts: 19,020
Blog Entries: 5
Default

Yes. The problem is similar to what I described, you may not want that heat concentrated in one place. Like I said, you may need to spread it (the heat) around, it is going to be there no matter what.

I'm more worried about what the heat will do to the surrounding circuitry, if it isn't that intense you need not worry about it. If it is think of ways to get rid of it, so it doesn't slowly bake anything to death.

And yes, you can up the wattages just fine. The leads will be heavier, but that is what they are for.
__________________
..
"Good enough is enemy of the best." An old engineering saying, Author unknown.

General info:
If you have a question, please start a thread/topic. I do not provide gratis assistance via PM nor E-mail, as that would violate the intent of this Board, which is sharing knowledge ... and deprives you of other knowledgeable input. Thanks for the verbage Wookie.
Reply With Quote
  #52  
Old 04-28-2012, 07:13 PM
elec_mech elec_mech is offline
Senior Member
 
Join Date: Nov 2008
Location: Virginia, USA
Posts: 1,260
Default

Hey Chris,

What are the specific part numbers of the digits you are using? You mention 3.3-3.8VDC, so I assume you've got either got green, blue, or white.

As far as the resistor values, what voltage are you working with to power the displays, 8" in your example? Be extremely leary of the values posted on these datasheets. They assume the minimum voltage of 3.3V, but as you stated, they could easily be up to 3.8V.

Let's do an example though. Say you're using an 8" display. According to the datasheet, the forward voltage of each segment is 23.1V (7 LEDs x 3.3V per LED). Now, this could be as high as 3.8V per LED and require 7 x 3.8V = 26.6V. Also remember that your ULN2003 can take 1-2V. So a 24VDC won't quite do it. You need some overhead to account for the current-limiting resistor and you don't want to tax the supply or have your LEDs appear weak.

For the same of example, let's say you get two 15VDC supplies rated 8A or more and tie them in series to get 30VDC at 8A. Now your resistor values will be 30-26.6-2= 1.4V -> 1.4V/150mA= 9.3Ω or less. Let's use 10Ω. To find the right power rating of the resistor: P = V*I -> 1.4V drop across the resistor * 150mA across the resistor = 0.21W, pretty close to 1/4W resistor. A 1/2W resistor would be good. Bill mentioned using two or more 1/4W resistors to spread out the heat. Remember two resistors in parallel with the same value will give you a total resistanace of half that value. So two 20Ω, 1/4W resistors connected in parallel will give you 10Ω with a 1/2W total - this assumes equal amounts of current go across each one which should be the case. Similarly, you could put three 30Ω, 1/4W resistors in parallel and get 10Ω at 3/4W, four 40Ω, 1/4W resistors in parallel will yield 10Ω at 1W, and so on.

Bill makes a great point that I would have missed - make sure you add some copper, i.e., make bigger traces on your PCB to the resistors and LED traces so they don't heat up too much. The bigger the trace, the less resistace across it, the less heat generated on the board.

For your 33V requirement, which could go to 38V, I'd suggest using two of the 24VDC supplies you found and putting them in series to give you 48VDC and 10A. Then worse case: 48-38V(drop across the LED segemnt, you'll need to check this with a meter)-2V(drop across the ULN2003 - check this too) = 8V -> 8V/120mA = 67Ω ~ 68Ω a common value. If you stick with one resistor per segment, then 8V * 120mA = 0.96W. So you'd want at least a 2W resistor. Of course, if you can get a smaller input voltage like 44V, then you'll drop your needed resistor wattage quite a bit.

I don't think there is a problem putting two power supplies in series, but I'd double check with Bill. I've done this with linear power supplies which are heavy and use a transformer, but I haven't tried this with switching power supplies like the one in your eBay link - this was probably what the guy at your electronics store was referring to. Just found this article, it looks pretty good and says you just need a couple of diodes: http://www.acopian.com/acopianPowerS...y.aspx?nsId=17

Again, I defer to Bill as this is outside my experience.

Hope this helps.
Reply With Quote
  #53  
Old 04-28-2012, 07:37 PM
Bill_Marsden's Avatar
Bill_Marsden Bill_Marsden is offline
Super Moderator
 
Join Date: Mar 2008
Location: Dallas, TX (GMT-5 w/ DST)
Posts: 19,020
Blog Entries: 5
Default

Just make sure the power supplies in question are floating. 24VDC should be enough though, I believe I can get your design down to that voltage with a little work. More on that later.

I had a different thought, though it is slightly off topic. We have a lot of requests for scoreboards here. I'm thinking of starting a series of articles for modules. There are many things in common on these units, but there are usually differences too.

If there were a source for these modules in the form of plans they could be combined to make whatever scoreboard is needed. It would have saved you a lot of work if it had existed.

I am not shy about using existing plans, and I'm looking at yours and mine. If I do this may I steal some of your work? Steal in this case meaning use with your permission, and credit given. OK, so steal is a bad choice of words. What do you think?

I haven't really looked at your LED setup at all. You should be able to get by with 24V with no hassle. LEDs are both fun and easy, if you follow the rules.

Mind if I check your display schematics out?
__________________
..
"Good enough is enemy of the best." An old engineering saying, Author unknown.

General info:
If you have a question, please start a thread/topic. I do not provide gratis assistance via PM nor E-mail, as that would violate the intent of this Board, which is sharing knowledge ... and deprives you of other knowledgeable input. Thanks for the verbage Wookie.
Reply With Quote
  #54  
Old 04-28-2012, 09:37 PM
chrischrischris chrischrischris is offline
Senior Member
 
Join Date: Feb 2012
Location: Melbourne, Australia
Posts: 261
Default

Hi Bill.

Modules - I absolutely agree. I'm more than happy for you to use my plans and hack them as you feel best. I was going to upload all finished work at the end anyway for anyone to use / alter. This work never the less is not really all mine. It's a combination of this fantastic forum (I couldn't have done it without you and elec_mech) along with websites "doctronics" and "Lauriers handy little hints" and "google".

At the end I'll give you a photo of each breadboard (all 4), a corresponding wiring diagram (4) a PCB printable layout (4), some notes on how I created the PCB, finished board photos and finally scoreboard jpg and mp4 showing it in action! (hope I get there).

As for you comment - floating power supplies. Not sure what that means.
Reply With Quote
  #55  
Old 04-28-2012, 10:29 PM
Bill_Marsden's Avatar
Bill_Marsden Bill_Marsden is offline
Super Moderator
 
Join Date: Mar 2008
Location: Dallas, TX (GMT-5 w/ DST)
Posts: 19,020
Blog Entries: 5
Default

Sometimes power supplies are attached to ground on the + or - leads, if the - side is grounded they are positive power supplies, and visa versa. You can not chain these together because of the ground. If they are not attached they are floating.
__________________
..
"Good enough is enemy of the best." An old engineering saying, Author unknown.

General info:
If you have a question, please start a thread/topic. I do not provide gratis assistance via PM nor E-mail, as that would violate the intent of this Board, which is sharing knowledge ... and deprives you of other knowledgeable input. Thanks for the verbage Wookie.
Reply With Quote
  #56  
Old 04-28-2012, 11:00 PM
chrischrischris chrischrischris is offline
Senior Member
 
Join Date: Feb 2012
Location: Melbourne, Australia
Posts: 261
Default

Hi elec_mech.

Sorry for the assumptions. I'm using ultra blue - wavelength 470.

I just tried to measure the voltage drops but came across a hurdle. Sorry for my lack of knowlege here, but the 8" display is common anode, hence each segment is -ve. Not sure how to measure the voltage drop. I just tried and killed one of the ULN2003's then a resistor (better that rather than the leds!). So I think I'll stop there.

Could you look at my circuit diagram from an earlier post and let me know how to measure the voltage drop considering the led displays are run via -ve to each segment.

I would however like to run these at a lower intensity to lengthen the life of the leds. I was thinking of between 15mA and 20mA rather than 30mA. Changing resistors, I found that the total current draw on all 7 combined segments was 700mA. The intensity of the LEDs was still quite bright - good enough (my kids were telling me to turn it off as it was so bright!). I then changed resistors to get it closer to 1amp - this was brighter than I need.

I would like to measure the voltages (as well as consumption of the ULN2003 - without blowing it up this time).

P=VI. Yes, I'll use this. As for the traces - I'll up those as well (or even use jumper wires.

Regarding the power supply link - nice link - ta.
I did read however (more a question for Bill), that in order to put 2 switching power supplies in series, you need to have an "isolated output". What does that mean?

Last edited by chrischrischris; 04-30-2012 at 11:58 AM.
Reply With Quote
  #57  
Old 04-29-2012, 08:46 AM
Bill_Marsden's Avatar
Bill_Marsden Bill_Marsden is offline
Super Moderator
 
Join Date: Mar 2008
Location: Dallas, TX (GMT-5 w/ DST)
Posts: 19,020
Blog Entries: 5
Default

I am guessing that is floating, though I am not sure.
__________________
..
"Good enough is enemy of the best." An old engineering saying, Author unknown.

General info:
If you have a question, please start a thread/topic. I do not provide gratis assistance via PM nor E-mail, as that would violate the intent of this Board, which is sharing knowledge ... and deprives you of other knowledgeable input. Thanks for the verbage Wookie.
Reply With Quote
  #58  
Old 04-29-2012, 07:18 PM
elec_mech elec_mech is offline
Senior Member
 
Join Date: Nov 2008
Location: Virginia, USA
Posts: 1,260
Default

Hey Chris,

How did you take the measurement to cause the ULN2003 to fail?

I looked at your last schematic from post #47 and noticed it shows your displays being tied to ground at pin 3. Since these are C.A. displays, is pin 3 actually tied to Vcc? I looked at the datasheet for the 8" display, but unfortunately they don't call out the pin numbers.

Assuming the common anode of each display is tied to Vcc, verify your meter probes are hooked into the meter properly. The red probe goes to V+ or similar on the meter and the black probe goes to V- or GND. Make darn sure your probes (usually red only) is NOT connected to the current measuring connection, usually located on the far left of the meter. This will short the probes and your circuit if not used properly to measure current.

Set the meter to read VDC (not VAC). Connect the red probe to the C.A. pin on one display (or simply Vcc going to the display). Place the black probe on one of the segment pins of the display (those pins going back to the resistors and ULN2003 pins). Make sure the resistors are connected between the ULN2003 and the display. You do not need to disassemble your circuit or short anything. Have a piece of paper handy and record the voltage for each segment, i.e., a, b, c, d, e, f, & g. This is the voltage going across each segment with your current resistor values. It will change a little bit if you change the resistor values. Also note you probably will see no voltage on a segment that is off, so try to set the display to show 8 to light up all segments or you can disconnect pin 3 of the corresponding 4511 IC from Vcc and connect it to GND. This will force all of the segments of that specific display to light up.

Now place the black probe on one of the outputs of the ULN2003. Connect the red probe to the corresponding resistor between that output and the display segment. This will tell you the voltage drop across the resistor. Do this for all seven outputs.

Now place the black probe on pin 8 of the ULN2003. Place the red probe on one of the 7 outputs. This is the voltage drop across the ULN2003. Do this for all seven.

Adding up the three voltage drops (across ULN2003, resistor, and segment) should give you Vcc feeding the display. If not, there is another drop somewhere that needs to be identified.

Knowing these values will allow you to better select or tweak your power supply or resistors accordingly. It would be a good idea to measure the current draw across each segment. To do this, move your red probe into the A (not mA) socket on the left of your meter (assuming you have this). Disconnect one end of one of the resistors, let's say the end connecting to the segment pin. Now connect the red probe to the corresponding segment pin and the black probe to the removed resistor lead. This will tell you the current going through each segment and resistor.

The current and voltage drop across each component will change everytime you use a different resistor, so you may need to repeat these measurements several times. Ideally, as you mentioned, you want to increase LED life by putting as little current through them as you need. Therefore, use your first measurements and current resistor values to see what value of resistors you can safely test as well as the needed resistor wattage you'll need. Keep increasing the resistor value until the display is as bright and/or dim as you want it.

Regarding your power supply, I think the one on eBay would work. It has a separate GND connection, so theoretically the COM (- output) connection is not tied to GND inside the supply. If you want to be extra safe, put a 500mA or 1A fast-blow fuse between the V+ and COM connections you hook together between two supplies.

So: Supply #1 Supply #2
COM V+ -----fuse-----COM V+

The unconnected COM of supply 1 is the GND of your circuit and the unconnected V+ goes to the common anode. Of course, assuming the fuse does not blow, test the output voltage first.

Once verified, you could connect the COM of supply #1 to the GND connection of the supply if desired for some protection, but that would not be required for operation. DO NOT connect the COM from supply #2 to GND of the supply. You should also use the continuity function of your meter to verify that COM is not connected to GND internally before doing anything.

When deciding on brightness, there are a couple of things to consider:
1) Distance the displays will be viewed from - if your display appears too bright up close, make sure you are viewing it from the same distance and angle as your target audience before changing the resistor values.
2) Daylight vs night - I think you already mentioned accounting for this, but just a reminder. A day time display will need to be brighter than a nighttime one.
3) Optical filter - Regardless of where or when the display is going/used, you want an optical filter of some type to better read the digits. Look at an LED bedroom alarm clock. They all have a colored plastic cover over the digits that match the color of the digits. This isn't to protect the LED so much as to make them easier to read. I suggest finding a 0.1 to 0.125 inch thick piece of transparent (dark) blue arcylic. As an example, something like this (the dark blue one): http://www.tapplastics.com/product/p...ent_colors/519.

I'm not sure what is available in your neck of the woods, but you should have a plastic sheet shop somewhere I'd imagine. The filter will help you see the digits much more clearly.
Reply With Quote
  #59  
Old 04-29-2012, 10:13 PM
chrischrischris chrischrischris is offline
Senior Member
 
Join Date: Feb 2012
Location: Melbourne, Australia
Posts: 261
Default

Hi elec_mech.
You're right on the ball again on a number of counts.

1. The schematic wiring diagram is incorrect. My breadboard and PCB layout has pin 3 of the displays going to Vcc. I didn't bother with Pin 8 - I assumed it was a double up as it worked without it being connected to Vcc. Correct?. As for the wiring diagram I forgot to change pin 3 to Vcc (as the digits were originally C.C types). Schematic amended. Ta.

2. I had been flipping back and forth the red probe to 10A (s I was also measuring at one stage the current between GND and the digit common (Vcc) to get a reading of between 600mA and 1.2A. I remember trying to measure between the C.A pin and the segment pins, so as you pointed out, my meter must have been incorrectly set. Your outline below makes perfect sense. I'll try it later today after work.

3. Power supplies - good to know how to set it up. I'll archive this info and use it down the track. At least I know it'll work. Another hurdle solved. Ta.

4. Brightness - you hit it right on the nose. I tried one display yesterday on the soccer ground in full sunshine (last time it was cloudy). I also walked across the other side of the pitch to check. Given the front of the displays are white, the contrast between the ultra blue and the white is not enough. The filter would probably work a treat. Also, with the current at the moment set to the lowest value, I assume the brightness is around 1400mcd (as per spec sheet) - 600mA total for 7 segments as compared to max of 1.05A. When I had it 1.05A however, I found the digit got really warm - I assumed this was not good. Not sure???

The spec sheet says I can run these at "typ" 4000mcd. I can't see the math? My 24V supply, via 10Ω resistors - total current (all 7 segments) = .6A, therefore probably min intensity 1400mcd. If I have a higher voltage, max total current for display of 1.05A, would this increase the intensity from 1400 to 4000 (i.e. factor of 3'sh)? Double the current - wouldn't it only double intensity? And if at 4000mcd, is it ok the digits get so warm? Spec sheet says 4000 is "typical" not "maximum".

5. Biggest problem now. Viewing the displays from across the pitch (50 metres), the 8" is too small. The 6" would be rediculous. What an oversight. I'm so depressed!!! I may have to look then at 20" displays throughout. Voltage required still 23 to 26V, however current per segment = 25amp x 18 = 0.45A. This is 3.15A per display. For 15 displays all lite up at once (not typical), it would be 15*3.15A = 47amps at 23V DC. This would be probably 3 or 4 lots of 15Amp supplies which isn't the issue - it's the circuitry. I'd probably have to abandon the ULN's and use something else - mosfets? I need a whisky.
Reply With Quote
  #60  
Old 04-30-2012, 12:42 AM
elec_mech elec_mech is offline
Senior Member
 
Join Date: Nov 2008
Location: Virginia, USA
Posts: 1,260
Default

Hey Chris,

Pin 8 of the display? If you have two pins for Vcc, connect them both to Vcc directly. Remember each segment can take up to 150mA and that's from Vcc to ground. If you only use one Vcc pin, that means you're pumping 7 x 150mA = 1.05A through one pin. If you use both, then you divide up the current flow by half or 525mA per pin.

On the subject of your schematic, I noticed earlier but forgot to ask about your timer selection wiring solution. Looks great and has given me a few ideas for future projects. One observation though, I notice you're feeding the Vcc input of the BCD switches from the 4017, but they're still connected to Vcc directly. I assume you only have them connected to the 4017, correct?

Quote:
2. I had been flipping back and forth the red probe to 10A (s I was also measuring at one stage the current between GND and the digit common (Vcc) to get a reading of between 600mA and 1.2A. I remember trying to measure between the C.A pin and the segment pins, so as you pointed out, my meter must have been incorrectly set.
I've done that a few times too. Took a while before I figured out I had blow the ammeter fuse on the meter when I couldn't get it to work right later.

Quote:
Given the front of the displays are white, the contrast between the ultra blue and the white is not enough. The filter would probably work a treat.
Yes, the difference is like night and day, no pun intended. As a real quick, cheap method, you could also pick up some blue Syran wrap from the grocery store, cut off a bunch of 1 ft sections and layer 3-4 of them. This would at least give you something to "see" until you found a blue acrylic source.

Quote:
Also, with the current at the moment set to the lowest value, I assume the brightness is around 1400mcd (as per spec sheet) - 600mA total for 7 segments as compared to max of 1.05A. When I had it 1.05A however, I found the digit got really warm - I assumed this was not good. Not sure???
It's kind of hard to say if the actual luminous output is a match to the spec sheet (1400mcd), the big factor here is how it looks to you and your requirements. Unfortunately, pre-built 7-segment displays get a little pricey the bigger you go and brighter ones add up more. To boot, they use small SMD LEDs that are then coated in epoxy to give you a uniform segment. That is great, but it diminishes the brightness. So first, tie all available Vcc pins on the displays to Vcc to disburse the current and see if that helps. Next, measure the actual current consumption of each segment to make sure it is fairly uniform. Of course, if they weren't one or more digits would appear dimmer than the others, but it never hurts to verify. Warm isn't unexpected, but they shouldn't get hot.

Quote:
The spec sheet says I can run these at "typ" 4000mcd. I can't see the math? My 24V supply, via 10Ω resistors - total current (all 7 segments) = .6A, therefore probably min intensity 1400mcd. If I have a higher voltage, max total current for display of 1.05A, would this increase the intensity from 1400 to 4000 (i.e. factor of 3'sh)? Double the current - wouldn't it only double intensity? And if at 4000mcd, is it ok the digits get so warm? Spec sheet says 4000 is "typical" not "maximum".
Oh, there is a bit about optics, luminous intensity, viewing angle, current consumption and so on I've read about, but I can't recall anything distinctly to help you here. Maybe Bill has some input? In any event, you're using a 24VDC supply for digits that require 23.1V? That is cutting it awefully close. I've suggest upping your input voltage a little to say 28-30VDC and adjust your resistors accordingly. What might be happening is you're reaching the max voltage limit of the supply and the display is pulling more current than it should to compensate. Note the datasheet says the input voltage could be as high as 26.6VDC. I know it says max, but it's always a good idea to provide, say 3-4 volts over the required voltage of the LEDs just to be safe. This gives you that extra "pressure" if you need it to drive them.

Quote:
5. Biggest problem now. Viewing the displays from across the pitch (50 metres), the 8" is too small. The 6" would be rediculous. What an oversight. I'm so depressed!!! I may have to look then at 20" displays throughout. Voltage required still 23 to 26V, however current per segment = 25amp x 18 = 0.45A. This is 3.15A per display. For 15 displays all lite up at once (not typical), it would be 15*3.15A = 47amps at 23V DC. This would be probably 3 or 4 lots of 15Amp supplies which isn't the issue - it's the circuitry. I'd probably have to abandon the ULN's and use something else - mosfets? I need a whisky.
Don't fret, let's say you need the digits bigger and/or brighter, you have another option - make them yourself. I've done this using 1/8" thick gray PVC plastic. This is great stuff because it's relatively cheap and it is super easy to machine (read: drill without splitting, which my first favorite material, acrylic, does easily). You then buy super bright, cheap LEDs from ledshoppe.com or eBay and wire up your own display. It will take some time to make, but it will be brighter and cheaper than any 7-segment display you'll find. To boot, you can wire each segment to the optimal voltage and current requirements of your circuit and power supply. Only difference will be small individual LEDs making up a segment instead of a uniform, diffused one (in appearance). Just a thought. You'll still want a filter though.

Try taking a poster board or tape a bunch of papers together and use a black marker to make a couple of different-sized "8" digits. Put them where you'd mount your display on a bright, sunny day and stand where your target audience will be. This will help you decide on what size you need.

Good luck!
Reply With Quote
Reply   Post New Thread

Tags
, , ,


Related Site Pages
Section Title
Worksheet Counters
Textbook Ring counters : Shift Registers
Textbook LED sequencer : Digital Integrated Circuits


Similar Threads
Thread Thread Starter Forum Replies Last Post
How do I make a specific MCU pin HIGH for 1hr. poxkix Embedded Systems and Microcontrollers 35 01-06-2012 04:31 AM
A Digital count down Timer Schematic dandeman360 The Projects Forum 4 11-24-2011 01:13 PM
LED count down timer brozizds The Projects Forum 38 11-12-2011 02:21 PM
TIMER to count for 1 second honeymustard909 The Projects Forum 1 08-30-2011 12:11 AM
Two stage timer. Need advice. Airasolt The Projects Forum 2 09-10-2009 12:15 PM

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

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT. The time now is 07:13 PM.


User-posted content, unless source quoted, is licensed under a Creative Commons Public Domain License.
Powered by vBulletin
Copyright ©2000 - 2014, vBulletin Solutions, Inc.