Good evening everyone. I am hoping for the best, but I will need your help with the best course of action.

I have been using the 5" Display circuit shown below for some time now. While working on replacing a different power supply (separate from this circuit), I quickly found out I screwed up.

When I powered everything up, my display did not come on. Hmm, that's strange I think to myself. Just as I step around the corner to have a look see, I see a puff of my first "magic smoke" and to top it off, I even get a small fire to boot!

The V+ wire for the 20V supply had come loose and was making contact with one of the 100 ohm resistors. Houston, we have a problem! I had some extra 150 ohm resistors handy, so I replaced the fried resistor.

I was obviously concerned about other damage, but all I could do was power it up. When I powered up, everything looked great - 3 zeroes (I was expecting 1 segment to be a bit dimmer, but I couldn't really tell.

But, after watching the displays count, I quickly found a problem with the 2 digit. Every time a 2 is displayed, the "c" segment also lights up (should just be a,b,g,e and d. All other digits display fine.

I am not sure the best way to attack the problem. Is it possible the 4553 or 4543 got messed up as well? Or should I replace all of the 7-segment resistors?

So, how bad is it?

 

Do you have an oscilloscope or only a DVM?

What is Q4? Is this a resistor pack?

My guess is that U3 4543 needs to be replaced.

You should be able to test this by probing U3 pin-11 (c output) when the number 222 is being displayed (if all you have is a voltmeter for testing).

 

I don't have a definitive answer to your problem, however I notice that the C segment is used in all numbers 0 through 9, except for 2. So, I speculate that something is triggering the C segment for all numbers, but because all except 2 need the C segment, that is the only time the error is evident. In other words, I would tend to believe that the C segment is permanently enabled and would look to find a reason why. Hope it helps.

 

Darn. From the title, I was expecting to see a picture of flames and a panicked geezer frantically swatting them with his hastily yanked off t-shirt

 

Do you have an oscilloscope or only a DVM?

What is Q4? Is this a resistor pack?

My guess is that U3 4543 needs to be replaced.

You should be able to test this by probing U3 pin-11 (c output) when the number 222 is being displayed (if all you have is a voltmeter for testing).

The Q4 is a ULN2804. I cropped my larger circuit which had all of the designations - sorry about that. I only have a DVM. I will give your suggested test a try tomorrow for sure.

Thanks for your help.

 

I don't have a definitive answer to your problem, however I notice that the C segment is used in all numbers 0 through 9, except for 2. So, I speculate that something is triggering the C segment for all numbers, but because all except 2 need the C segment, that is the only time the error is evident. In other words, I would tend to believe that the C segment is permanently enabled and would look to find a reason why. Hope it helps.

Tracecom - good catch about the c segment. I didn't realize that. Duh! I should have noted which segment resistor got fried - I will have to check tomorrow - but smart money is c.

 

In that case, replace Q4 ULN2804 first.

(You are using only 7 channels of 8 on the package. You can try switching "c" to the spare channel.)

 

In that case, replace Q4 ULN2804 first.

(You are using only 7 channels of 8 on the package. You can try switching "c" to the spare channel.)

MrChips - Nice call on the extra channel. Switched segment c to the extra channel and I am back in business. Nothing like a little destructive test to confirm that 20V will kill a ULN2804 channel! Glad I could contribute ... !

Thanks everyone. I was fortunate it wasn't worse.

 

Why use 20v? It is high, and will likely damage your LED displays with excess reverse voltage as the matrixing exposes the off LEDs to -20v.