Shot clock LED troubles

Hello. I've been taking care of a shot clock for a local community
college, and I need some advice.
The clock has 4 LEDs for each line, with two full digits (7 lines per
digit). They are surface-soldered to a large, custom circuit board.

The problem I have is that the lights very often will go out halfway
or completely. Since each line is in series, a single bad LED will
knock out a whole line. An LED that is having trouble sometimes causes
the whole line to flash or flicker.

I check the contacts and the LEDs and they are all working perfectly
up until the date of use. Does anyone know why this would be?
ex: I get it home, I replace any bad LEDs, and all of them work. I'll
check on it periodically until it is needed for a game. The day that
it is needed, LEDs go out almost immediately after it is started up.

The LEDs actually go bad. The contacts usually are not damaged enough
by movement and such to cause this issue. I'll get it home the next
day and I'll find LEDs on it that just don't work anymore.

Are LEDs themselves sensitive to jostling? I try not to damage them
while soldering. Could it be a power cleanliness issue perhaps?

Any suggestions would be appreciated.

Thanks

 

I would check if the LEDs are being over-driven.

 

The forward voltage ranges from 1.9-2.1V on the contacts, and the LEDs
are rated as optimal at 2.0 to 2.5.

 

I don't think that is what Jumpster meant.
We need to know how much current is being sent through that LED. How
many mA are being consumed and what is the rating for each LED. Use a
simple ammeter (or milli-ammeter in series with one of the LEDs.
A much better design would be to run each LED in parallel with each LED
having its own current limiting resistor. This would prevent all LEDs
from going out at the same time.

I guess a schematic would be nice too.

 

no schematic but basically, it's a shot clock with multiple
independent LED drivers (one per "line", with 7 "lines" per digit).
Each "line"/driver has a 5ohm resistor in series, along with 6 LEDs in
series, and a total of 12V cumulative across the series circuit (I
measured 2V per LED). I have no information on the specs/tolerances of
the LEDs that came with the shot clock (the manufacturer refuses to
disclose this) but my replacements are wp1503id available at
mouser.com, data sheet at http://www.us.kingbright.com/images/catalog/
SPEC/WP1503ID.pdf
I measured 45mA after opening a spot in the circuit, and since its in
series that should be the same across all components in said circuit,
correct?

The specs do seem to show that the standard current should be 20mA
(2V-2.5V at 20mA) with an absolute max of 30, so I guess thats the
culprit? Should I just boost the resistors or am I being a dumbass in
that assumption?
I'd love to have the whole thing be in parallel but I don't have that
kind of freedom with this thing... at least not without making it into
a mess...

What would be the suggestion here then?

Thanks for all the help guys I unfortunately have no formal
electrical experience (that much should be obvious by now) and I do
appreciate the help.

 

If the LED string is being multiplex driven, then it's hard to measure the
true peak current that is being driven through the LEDs. You would have to
measure the current by measuring the voltage across the series limiter
resistor with a 'scope, and then doing the math. That aside, if you are
measuring 45mA with a DC ammeter, then it's probably at least that, which
seems wildly excessive - even for the original LEDs. You are correct in that
' rule of thumb ' for years ago was 10 - 20 mA. Typically, a single panel
indicator LED was driven from a convenient 12 or 15v rail, with a 1k ohm
series resistor, giving a LED current of 12 or 15 mA. Modern LEDs are much
more efficient than this, and my rule of thumb these days is about 5 to 10
mA for the same light output as an ' old ' type. They are still quite happy
for the most part, however, with up to 20 mA.

Nothing dumbass about your suggestion of upping the resistor values - it's
what the math would suggest, but I would have to question what is the cause
of you having to. Just over 2v per LED is normal for a red type, so string 6
of them in series, and you need 12v or so, to run them - ergo, no series R
required at all. This is a common scheme. A resistor is usually included for
safety reasons, to limit the current in the event of one or multiple short
circuit failures of the LEDs, or a physical short on the string. Under
normal circumstances, the voltage drop across this very low value resistor -
5 ohms in your case - will be minimal, so the power dissipation in it will
be low, and it will run totally cool. You could try tripling the value of
the resistor to 15 ohms, and measuring again, just to see what happens. If
you read a lower current, and the brightnes of the LEDs is still acceptable,
then this may be enough to ' cure ' the problem, but might be doing it by
masking an actual fault.

Without seeing a schematic for the whole thing, to see just how the LEDs are
driven - DC, or pulsed and multiplexed - upping the resistor is about the
most valid thing that I can suggest, which might result in a long term fix.
It is possible that there is a genuine intermittent problem with the drive
electronics, but it's curious as to just why this only seems to show when
the thing is in service, rather than at your home where you are fixing it.
Two things come to mind here. The first is temperature. Some intermittents
are particularly temperature sensitive. The second is line supply stability,
although if the thing is designed properly, it should have fully stabilized
supplies, and should not care too much about line power fluctuations,
provided that they are not wildly excessive.

Arfa

 

Or the voltage regulator may be faulty or intermittent.

--- sam | Sci.Electronics.Repair FAQ: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Sites: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: Anything sent to the email address in the message header above is
ignored unless my full name AND either lasers or electronics is included in the
subject line. Or, you can contact me via the Feedback Form in the FAQs.

 

If you replace one OEM LED in a string, you should probably replace the whole
string because you have no idea of the ratings of the originals.

Also, you might try paralleling each LED with a fixed resistor (all the same
value for that string of LEDs. You might aim for 10 ma through the resistor
string, that would be 2 volts/10 ma or 44.4 ohms for each resistor. It WOULD
increase the load on the LED drivers by about 11%, which might be a problem.)
to 'equalize' the voltage drops across each diode in the string of LEDs.

How close to 2 volts are you getting? If, for example you have one at 2.5 and
another at 1.5, differences might allow one LED to overheat and go down in
resistance and efficiency, over heating more and overloading other LEDs in
the same string until one or more fail.

The same 'through' each component, YES.
Current goes through. Voltage is measured ACROSS.



--
bz 73 de N5BZ k

please pardon my infinite ignorance, the set-of-things-I-do-not-know is an
infinite set.

remove ch100-5 to avoid spam trap

 

Why? LEDs are current driven. Their exact voltage drop is rather irrelevant
as long as it isn't far off.

Sorry, if they were that different, one of the LEDs is bad. LEDs of a given
color from a given manufacturer will be quite close in voltage at the same
current.

--- sam | Sci.Electronics.Repair FAQ: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Sites: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: Anything sent to the email address in the message header above is
ignored unless my full name AND either lasers or electronics is included in the
subject line. Or, you can contact me via the Feedback Form in the FAQs.

 

Holy crap, yeah 45mA is way too much if those are ordinary LEDs. Replace
the resistors with some of higher value to obtain about 18mA, if you
measure the forward drop across the whole string of LEDs, you can use
Ohms law to easily calculate the resistor value.

 

Agreed, As long as it isn't far off. I do know that strings of diodes often
need equalizing resistors but that may be to make sure that the reverse
voltage doesn't 'pile up across one diode and blow it'. I also seem to
recall that there are times when zeners in series require parallel
resistors.

Maybe he needs some capacitors rather than resistors.
Perhaps transients are killing the LEDs. Something is.

The fact that it works fine in the shop is puzzling.

With some such problems I use a variac to run the voltage up to 125 or 130
to 'stress test' it.

No need to be sorry.

Hence my suggestion that he replace the whole string when one goes bad.





--
bz 73 de N5BZ k

please pardon my infinite ignorance, the set-of-things-I-do-not-know is an
infinite set.

remove ch100-5 to avoid spam trap