1 Second Clock with 5" LED Displays Circuit Evaluation

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
elec mech to my rescue ....

I think your power supply is a bit low. Assume worst case scenario, all segments lit, 7 segments x (let's say) 20mA x 3 digits = 420mA. Now, I could be off on this as the datasheet seems to infer each strand of LEDs takes 25mA, but each segment would then consume 50mA. If this is correct, then we're talking 7 x 40mA x 3 = 840mA. I'm hoping Bill will chime in with his thoughts regarding what the true current per segment should be.
I will have to have some guidance on measuring the current. I def. do not want to short something out! I also noted some of the measurement you had posted when using the same displays, but the power supply may have been different in your case.

So, what exactly is happening when you connect the 5" displays to your circuit? Are they dim? Are only certain segments lighting up? Are certain segments brighter than others? Do all the digits appear to be off?
When connected to the circuit, all 3 displays appear off (possibly extremely dim, but to the eye they appear off. I have the 24V supply hooked up at the 3 2907's and the ground connected to the 9V supply ground.

When I remove the 24V ground wire from the circuit and touch it to each individual segment (after the 2804), all 3 corresponding segments come on - touch seg. a - all 3 a's light up., etc ...

What is the voltage across the C.A. pin and one of the segment inputs when your circuit is on? Do this for several segments and for all digits.

What is the current going into each segment? Same as before, please do this for multiple segments for each digit. Let us know if you're unsure how to do this with your meter and I'll explain. If there is any doubt, please ask first because it is possible to short something out if this is done wrong.
I will check voltages, but as I mentioned above, I will need some help on the proper way to measure the current.

Thanks again for jumping in to offfer help ...
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
What is the voltage across the C.A. pin and one of the segment inputs when your circuit is on? Do this for several segments and for all digits.
Let me know if this is what you were looking for regarding the voltages. I placed the meter ground at ground on the breadboard and then tested the back of the displays at each pin with the meter positive ...??? Is that the correct method, or are you looking for something else?
Display 1
CA - 20.2V
Segments: 11.81V, 12.04V, 11.99V, 11.97V, 11.98V, 11.92 V, 11.87V

Display 2
CA - 19.6V
Segements: 11.88, 12.05, 12.17, 12.07, 11.94, 11.99, 12.04

Display 3
CA - 20.9V
Segments: 11.89, 12.04, 12.07, 12.04, 12.06, 11.99, 11.97

A quick separate question - if the power supply is lacking, could I just try to power up 1 display only to check that? Or is the supply not even enough for 1 display?
 

elec_mech

Joined Nov 12, 2008
1,500
Hey CoachKalk,

Hmm, interesting. It seems you're seeing a 12V drop across the resistors and the 2804 to ground and only 8V across the LED segments.

Please take the following measurements:
1) Black probe on ground, red probe on Vcc2 (24VDC input).

2) Black probe on CA, red probe on Vcc2 input to corresponding 2907. Repeat for all three.

3) Black probe on segment pin, red probe on CA (both of same digit). Take multiple readings.

4) Black probe on one side of resistor connected to 2804 output, red probe on other side of same resistor. Repeat for all seven.

5) Black probe on ground, red probe on 2804 output pin (before resistor). Repeat for all seven.

6) Take current reading of 24VDC supply. To do this,
a) Unplug the 24VDC supply from the wall.
b) Plug the red probe into the 10A (or 20A depending on meter) socket of the meter - this is usually on the far left side.
c) Remove the positive lead of your 24VDC supply from your circuit.
d) Connect the red probe to this positive lead (if possible, use a probe with an alligator clamp instead of a point or connect the two using a separate wire with alligator leads). Make sure any exposed contacts are not touching anything such as a metal tabletop, part of your circuit, etc.
e) Connect the black probe of your meter to the Vcc2 input of your circuit. Again, use alligator clips if possible.
f) When taking a current measurement, you are putting your meter in series with the power you are trying to measure. If you think of current as flowing from positive to negative, you always want to make sure the red probe is connected to the positive-most side of the power path in question and the black goes to the negative-most side.
g) Set the meter to measure current at its highest setting, usually 10A (or 20A).
h) Plug the 24VDC supply back into the wall and see what the current measurement is.
i) Be careful to reconnect the red probe back into the resistance/voltage socket (usually on the far right of the meter) when you go back to measure voltages, otherwise you'll cause a short.

Yes, go ahead and try hooking up only one digit and leave the other two unconnected. If you don't mind, repeat the steps above once you've done this and we can see if there is any difference.
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
Hey CoachKalk,

Hmm, interesting. It seems you're seeing a 12V drop across the resistors and the 2804 to ground and only 8V across the LED segments.

Please take the following measurements:
1) Black probe on ground, red probe on Vcc2 (24VDC input).
21.97V

2) Black probe on CA, red probe on Vcc2 input to corresponding 2907. Repeat for all three.
Unless I screwed something up, I couldn't get any measurable voltage. It wasn't zeros, but very very close.

3) Black probe on segment pin, red probe on CA (both of same digit). Take multiple readings.
Again, no luck measuring any voltage on any of the 3 displays to any of the 7 segs.

4) Black probe on one side of resistor connected to 2804 output, red probe on other side of same resistor. Repeat for all seven.
At this point, I was convinced I had something hooked up wrong - no voltage across the resistors, but I just double checked the Vcc voltage and it measured same as before - 21.97V.

5) Black probe on ground, red probe on 2804 output pin (before resistor). Repeat for all seven.
a = 11.88, b = 11.80, c = 11.90, d = 11.73, e = 11.70, f = 12.34, g = 12.82

At this point I just started checking different areas. I am not sure if it helps, but I want to post these numbers. The values above were measure just as you asked - black probe:ground and red probe:2804 output, before resistors.

I then placed the red probe at Vcc (24V) and the black probe on the INPUT side of the 2804.
a = 22.12, b = 22.11, c = 22.06, d = 22.04, e = 22.00, f = 22.04, g = 22.05

6) Take current reading of 24VDC supply. To do this,
a) Unplug the 24VDC supply from the wall.
b) Plug the red probe into the 10A (or 20A depending on meter) socket of the meter - this is usually on the far left side.
c) Remove the positive lead of your 24VDC supply from your circuit.
d) Connect the red probe to this positive lead (if possible, use a probe with an alligator clamp instead of a point or connect the two using a separate wire with alligator leads). Make sure any exposed contacts are not touching anything such as a metal tabletop, part of your circuit, etc.
e) Connect the black probe of your meter to the Vcc2 input of your circuit. Again, use alligator clips if possible.
f) When taking a current measurement, you are putting your meter in series with the power you are trying to measure. If you think of current as flowing from positive to negative, you always want to make sure the red probe is connected to the positive-most side of the power path in question and the black goes to the negative-most side.
g) Set the meter to measure current at its highest setting, usually 10A (or 20A).
h) Plug the 24VDC supply back into the wall and see what the current measurement is.
i) Be careful to reconnect the red probe back into the resistance/voltage socket (usually on the far right of the meter) when you go back to measure voltages, otherwise you'll cause a short.
I triple checked your instructions here and had no luck. The current readings were not zero, but in the low mA range 10 mA. I have to head out to my daughter's softball game, but I will try the current read again this evening with a fresh set of eyes.

Yes, go ahead and try hooking up only one digit and leave the other two unconnected. If you don't mind, repeat the steps above once you've done this and we can see if there is any difference.
Tried and failed. 1 display only - still lights out.
 

elec_mech

Joined Nov 12, 2008
1,500
What is the measurement for 1) if you completely remove all the digits from the circuit?

Remove the 24VDC supply from your circuit completely. What is the voltage across the + and - leads?

Can you try putting in smaller digits to verify the circuit as a whole is working? If yes, please take the previous measurements 1-6 and post your results.

I'm not sure what is going on yet, but we'll figure it out.
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
My daughter caught a ride to the game, so I had a few extra minutes so I tried the following.

Supply Ground - to - Display1SegmentA - to Display1 CA - to - Tester(Black Probe) - METER - Tester(Red Probe) - to - Vcc.

Only 1 display
a segment only: 1.53 mA
added a and b: 2.96 mA
a+b+c: 4.41 mA
a+b+c+d: 5.82 mA

Added 2 display
a+b+c+d: 11.05mA

Added all 3 displays
a+b+c+d: 15.46 mA

Note: Because I was trying to rule out the supply source, I used a 20V, 2A supply I had. I double checked the voltage after running these tests and it measured 20.62V.

I quickly hooked the "new" supply to the circuit, but the displays still remained off.

Suggestions?
 

elec_mech

Joined Nov 12, 2008
1,500
So you connected the - 20VDC supply wire directly to segment a, then, through the meter, connected CA to the + 20VDC supply wire?

Then, you jumped segment b to a, then c to b to a, then a+b+c+d. Is this correct?

Did the segments appear lighted? Were they bright?

Can you repeat this test, except leave out the meter (no need to measure current since we know what it will be) then test the voltage from the segments to CA (red probe on one of the ? What we're looking for is whether the voltage is dropping significantly. Because you said the 24VDC supply is only putting out 21.97VDC, it infers the digits are pulling more current than the supply can provide. Whenever the current draw is more than the supply can provide, the voltage of the supply will drop below its rated value.

What is the voltage of the 24VDC supply when it is not connected to the circuit?

Again, do you have smaller digits you were working with before to test the circuit and 24VDC supply (assuming you were using this same supply previously)?
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
So you connected the - 20VDC supply wire directly to segment a, then, through the meter, connected CA to the + 20VDC supply wire?

Then, you jumped segment b to a, then c to b to a, then a+b+c+d. Is this correct?

Did the segments appear lighted? Were they bright?

Can you repeat this test, except leave out the meter (no need to measure current since we know what it will be) then test the voltage from the segments to CA (red probe on one of the ? What we're looking for is whether the voltage is dropping significantly. Because you said the 24VDC supply is only putting out 21.97VDC, it infers the digits are pulling more current than the supply can provide. Whenever the current draw is more than the supply can provide, the voltage of the supply will drop below its rated value.

What is the voltage of the 24VDC supply when it is not connected to the circuit?

Again, do you have smaller digits you were working with before to test the circuit and 24VDC supply (assuming you were using this same supply previously)?
As I was driving to the softball game, I realized my post may not have been clear. The current readings were measured as you posted - all of the other circuit had been removed. Simply Supply Ground to Supply Voltage with only the LED displays in play. The brightness was very very good!


I probably need to clarify the power source issue for the 5" LED's. I have two options available.

Option 1 - the supply has written on it: Output: 24V - 400 mA
Option 2 - the supply has written on it: Output: 20V - 2A

I started testing the circuit with Option 1. All of the measurements I had been providing had been with Option 1. When I struggled to get any current reading using the method you outlined, I began wondering if something was wrong with the supply. I then looked in my box of goodies and found/remembered Option 2.

So, the "current" measurements I had posted were actually with Option 2.

Supply Info
Option 1: 24V - 400mA
No circuit, just testing across lines: 21.17V
3 Displays with 4 segments together: 16.21V

Option 2: 20V - 2A
No circuit, just testing across lines: 20.62V
3 Displays with 4 segments together: 20.29V

Again, do you have smaller digits you were working with before to test the circuit and 24VDC supply (assuming you were using this same supply previously)?
I had the exact same circuit in place except I was using 1" displays. Obviously the 5" displays necessitated the extra action coming out of the DS pins on the 4553.

I "plugged" my 1" 3-display unit into the circuit we are discussing now, BUT I only used the 9V supply. I figured it would let me know if the 4553/4543/2804 portion was correct. I could not get the smaller display to light up at all.

Unless something jumps out at you, I may just manually check the LED displays just to make sure I didn't fry them and then start the circuit over to make sure I do not have a stupid wiring mistake.

I am quickly remembering the frustation from last Fal l
!!!
 

elec_mech

Joined Nov 12, 2008
1,500
Supply Info
Option 1: 24V - 400mA
No circuit, just testing across lines: 21.17V
3 Displays with 4 segments together: 16.21V

Option 2: 20V - 2A
No circuit, just testing across lines: 20.62V
3 Displays with 4 segments together: 20.29V
This tells us a lot. The 24VDC supply is suspect since the actual output voltage is well below the rated one. That aside, you notice the voltage drops by 5V when you connect it to your circuit? That means the circuit is drawing more current than your supply can provide. This either means there is a problem in your circuit causing more power to be drawn than it should be or that the circuit is fine and the power supply isn't big enough. I suspect the latter.

As you mentioned, check each segment of each 1" digit beforehand using 9VDC and a current-limiting resistor (9-3.7 = 5.3 / 0.025 = 212Ω) to be sure they work.

If you're testing the 1" digits, I assume they are rated for about 3.7VDC forward voltage and up to 25mA? If so, remove the zener diodes from the DS pins on the 4553 as well as the 330Ω resistors. Replace the 240Ω base resistors with something like 560Ω, limiting the current through the transistor to about 160mA (160/7 = 23mA per segment). Assuming I've learned something from SgtWookie on transistors anyway. Feed 9VDC to Vcc2. Do the 1" digits work now?

Next, let's see what the 5" digits need. Test each segment with your 20VDC supply and a current-limiting resistor (20.2-18 = 2.2 / 0.025 = ~100Ω).

How bright are the segments (only test one at a time)?

Repeat using a ~50Ω resistor. They will appear brighter.

Now, do the segments appear bright or dim with the 100Ω?

Do the segments appear bright enough or too bright with the 50Ω?

We're trying to narrow in on the spec sheet to find out if each segment needs 25mA or 50mA for normal operation. I suspect the latter, but only a test will show it.
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
First - the UGH!!! Before I even started the 5" project, as you remembered, I had worked on a 1" project. I had some issues with bounce, but the display lit and responded. Fast forward to my issues with the new 5" circuit. After the troubles, I tried the smaller displays in the circuit for the 5" displays - NOTHING!!!

The insane frustration came when I tried to re-hook up my previously working 1" display - even on the same circuit that had worked - NOTHING!!!! I tried to think through what I did and the only thing I could remember was borrowing 1 of the 2n2907's from my "small" set-up to use in place of one in the "big" set-up. That may have been when I was grasping I think.

Last night after my small circuit would't work, I swapped out the 4543 and 2804 for new IC's - just in case. Still nothing! This morning I continued to check components and found the transistor check - I tested all 6. BAM! 1 of the 6 tested bad???!!!! Problem solved I think. I make sure I have 3 good 2907's in play and fire up the 1" circuit ......... NOTHING AGAIN!!!! ARE YOU SERIOUS????? The only other option I can do is change out the only IC that I hadn't got to yet - the 4553.

Power it up and WINNER WINNER CHICKEN DINNER!!!!!!!!!!!!!!!!!!!!!

So, finally I had some good news. My old circuit was back at the party. I say to myself, "Self, maybe you should put the new 4553 in the "big" circuit that was converted for the 1" trial (removed diodes, 560 Ohm base resistors. DOUBLE BAM! 1" displays working.

Now, I am crazy nrevous about setting up the official "big" set-up. I have no idea what caused what. A few questions I have:
1. Would/could a bad 2907 damage the 4553?
2. Or, would/could a bad 4553 damage the 2907?
3. If my 24V supply has been whacked from the beginning, could that be to blame for all failed components?

Regardless, I did complete the tasks below and would like to provide as much info as possible.

As you mentioned, check each segment of each 1" digit beforehand using 9VDC and a current-limiting resistor (9-3.7 = 5.3 / 0.025 = 212Ω) to be sure they work.
Good to go, tested great with no other components - yeah, I didn't kill them!

If you're testing the 1" digits, I assume they are rated for about 3.7VDC forward voltage and up to 25mA? If so, remove the zener diodes from the DS pins on the 4553 as well as the 330Ω resistors. Replace the 240Ω base resistors with something like 560Ω, limiting the current through the transistor to about 160mA (160/7 = 23mA per segment). Assuming I've learned something from SgtWookie on transistors anyway. Feed 9VDC to Vcc2. Do the 1" digits work now?
After ensuring all 2907's were good AND replacing the 4553, everything worked.

Now, do the segments appear bright or dim with the 100Ω?.
5" displays - 20V supply, 100 Ohm: I would consider the brightness very acceptable. Per you instructions I only tested 1 seg at a time so I am not sure if the additon of more/all will cause the brightness to drop, but 1 at a time was very good.

Do the segments appear bright enough or too bright with the 50Ω? .
5" displays - 20V supply, 50 Ohm: Surprisingly similar to the previou test. Brighter for sure, but not crazy, looks like it will ignite bright.

So, with my comments above about the failed components, the possible issue with the 24V supply, my million dollar question - how to proceed with the 5" circuit? Is there something involving the components I had removed (resistor, 2 zener diodes) per DS pin that could have caused the failures? Do I need to double/triple check a key area before powering it up again? Obviously I do not want to fry anything else if I did do something wrong.

Thanks ...
 

elec_mech

Joined Nov 12, 2008
1,500
Hey CoachKalk,

First, ditch the 24VDC. Something is wrong with it if the output is ~21VDC when it is connected to nothing. What happened? Hard to say. It's possible the supply was bad to begin with and could have damaged the other parts, the load simply could have been too much and damaged the supply, static electricity could have gotten to either the IC or the transistor. If you're in a humid area right now, there's probably no need to fuss with it right now, but it's always a good idea to wear an anti-static strap connected to a good ground anytime you are touching ICs, transistors, etc.

Alrighty, so the circuit works, that's fantastic!

Let's step through this and cover all our bases. Repeat the tests on the 5" displays with the 100Ω and 50Ω resistor and take current measurements for each this time. Remove the meter and repeat the tests and take voltage measurements: across + and - of the 20VDC supply, across CA and the segment pin, and across the resistor. Presumably, the voltage across the + and - leads will be about 20.2VDC and the current will be ~20mA and ~40mA for 100Ω and 50Ω, respectively.

If that is not the case, stop here.

If it is the case . . .

Going with SgtWookie's math, the PNP will start to turn on at Vcc2-0.7. Let's assume the 20VDC puts out 20.2VDC under load, therefore 20.2-0.7 = 19.5VDC. Let's also shoot for about 23mA across each segment, so 23 x 7 = 161mA. We need 1/10 of that to feed the base of the transistor, so ~16mA. Now, 19.5V - 17.1V = 2.4VDC, 2.4 / 16mA = 150Ω for the transistor base.

Reconnect both Zener diodes for each of the DS outputs on the 4553. Replace the 560Ω with the 150Ω (previously 240Ω) and add the 330Ω resistor across each transistor again as shown in your schematic. There will be a small drop in voltage across the transistor and the 2804, but let's keep things simple for now and assume the voltage going through the segments is 20VDC. Each segment requires 18VDC according to the datasheet, therefore 20-18 = 2 / 23mA = ~87Ω. Try 100Ω resistors for R12-18 (resistors between segments and 2804).

Hopefully you should be able to power up the circuit and "see" something. If anything, the segments may appear dim. If that is the case you could try dropping the 100Ω resistors to the segments down to 50Ω. If that doesn't brighten them up enough, you may need a 24VDC supply (rated at least 2A).

I can't promise I didn't miss something, but your bases should be covered and hopefully won't result in any more damaged parts.
 

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
elec mech - I can not thank you enough for your help getting these 5" displays to work. I performed the "pre-tests" you recommended and the results were very close to the expected values.

I prepared the circuit for the 20V supply and was actually nervous when I plugged the supplies into the wall. BINGO! Even with the 100 Ohm resistors, the segments actually show up very well. I have not put together a cover for the diaplys yet, but unless the cover really takes away from the brightness, I am very pleased!

Thank you very very much!!!
 

elec_mech

Joined Nov 12, 2008
1,500
That is awesome Coach. Glad to hear everything is working.

I've attached couple of pictures showing a 4" display with and without a red acrylic filter. In both pictures, "00" is lit. The filter makes a huge difference by making the display easier to see. If you have a place locally that sells acrylic, you should be able to find some dark red transparent acrylic. If not, I get mine from Tap Plastics: http://www.tapplastics.com/product/...plastic/acrylic_sheets_transparent_colors/519. I use 1/8" thick.

I'm working on a project now and plan to drill two holes through the centers of each digit (picture a hole in the center of the "circles" of an 8), then countersink a hole in the acrylic. From there I'll use black flathead screws from McMaster to make the screws discreet. I've also designed a bracket made from aluminum that allows you to cut a rectangular hole in your mounting surface, say 1/4" plywood, and drop it in, then secure with screws. I can send some pictures and plans if you're interested.

If you can tell me how many digits altogether you need to mount, in what groups, rough positions, and what you're mounting them to, I can make some suggestions and come up with some plans to help.
 

Attachments

Thread Starter

CoachKalk

Joined Sep 20, 2011
141
elec mech - I am totally interested in any plans you have developed. Even if I don't use 100% of your design, I would love some assistance.

I checked out the website you posted - would you suggest the darker red versus the "hot" red?

Thanks again for all of your help!
 

elec_mech

Joined Nov 12, 2008
1,500
would you suggest the darker red versus the "hot" red?
Definitely the dark red. The "hot" red is pink and very light in color. The filter should be as close a match to the LED color as reasonably possible so it allows that color to pass. It is also "hiding" the unlit segments so that the lit ones are nice and clear.

If you look at a seven segment display without a filter it is hard to make out which segments are lit and which are not in a bright room, especially with fluorescent lighting. Take a look at LED alarm clocks - they all have a colored filter over the segments. You can't see the unlit segments unless you're close and staring right into the display. Now look at a seven segment display, say the 1" ones you've used. Even off, you can easily make out every segment and see an 8.

I've attached my current design for a mechanical assembly. I'm using an aluminum T-channel as the frame. I size the acrylic so it is the size of the combined digits plus an inch for each side. So if the digits are about 6" x 8", my acrylic will be 7" x 9". This accounts for the lip inside the frame based on the frame size (you may need to increase this if you use a wider T-channel - that is, if you opt to use this design :)).

The T-channel is cut at 45° angles to make a frame that will fit the acrylic, plus maybe a 1/16" on each side to allow a little play. The frame is laid out then the acrylic is set into it. A 1/4" aluminum bar cut length-wise is placed atop the acrylic inside the frame at the top and bottom. Holes are drilled and tapped on the ends of the bars as well as the top for the top bar and the bottom for the bottom bar (not shown in the attachment) and through the T-channel. The T-channel holes are then countersunk using a 100° countersink bit. 4-40, 100° flathead screws are then fastened to secure the T-channel, acrylic, and bars together. The thickness of the T-channel is 0.050" and the 100° allow for more/better gripping and a lower profile so the heads don't stick up above the T-channel.

This then allows a simple rectangular hole to be cut out in the mounting surface and the frame can be "dropped" in. From there, holes are drilled through the frame and mounting surface. Again, 100° countersinks are drilled in the T-channel and 4-40, 100° flathead screws are used to secure the frame to the mounting surface using lock nuts.

The digits themselves are centered in the acrylic. Holes are drilled through the acrylic and the two centers of each "8". The acrylic holes are then countersunk with an 82° countersink. 6-32, 82° black, flathead screws are used with lock nuts to fasten the digits to the acrylic. The black screws should hide nicely in the dark red acrylic.

I originally wanted to secure the acrylic directly to the frame as I did from the frame to the mounting surface, that that would then make the frame appear to look like swiss cheese with screws right next to each other. Instead I'm using the 1/4" bar to hold the acrylic to the frame.

I'm ordering parts now and hope to test my design in the next couple of weeks. I can take pictures and post them if you'd like.

If you'd like to do something else and need some assistance or advice, just let me know.
 

Attachments

Top