MB-LRE8x2CS need your help for something

[Lamouette]

Hi there, it is me again !

I am on the end of my 32 encoders board, everything works as expected  but there is just one thing, one stupid led which not work !

I explain :

When i turn an other encoder i can see a little bit of little across the ledring with the dead led. for each led lit, all the other reduce their brightness. it is the only one ring which do this, all the other one works as expected !

I checked the continuity between the 2 legs of the legs, and effectively they are linked together. so the led don't work, i checked all around, cannot find any place with a link between 2 solder, any other place i could chec or maybe it is another component dead ?

 

i already unsoldered 3 times, i started to scratch some part on the pcb, because i thought the problem was on the pcb itself.....

 

Any help ? Thanks !

 

Lamouette

 

 

[latigid on]

Perhaps try to swap around the shift registers to see if the problem follows. There shouldn't be a short between the LED pins except when the matrix is being driven. Also check the associated resistors. 

[Lamouette]

i just swap the ic, but nothing changed, i really think it is because of the short on the led, but i cannot find any point, and the layout of the project page don't fit with my PCB, so it is hard to follow the trace.

http://jeromebo.free.fr/Wiki/Layout.pdf

[latigid on]

Did you try replacing the LED?

[Lamouette]

yes, but as the link is between the 2 led pads, i don't expect it to work corectly, i seen that the pcb designer (fairlightIII) was not anymore on the forum, so i think it is dead to as k him the correct layout of the pcb...

[latigid on]

Are you using the Darlington drivers? Are the shorted pads connected to ground or any power rail? 

The schematic might also help: http://jeromebo.free.fr/Wiki/Schematic.pdf

For me, it's quite unusual that only one LED of a matrix is failing; usually it's a whole row or column. As a last resort you could cut the traces and wire manually, considering that this may break more connections in the matrix.

[Lamouette]

Yep, it is what i did, i just killed the pad, the remove the short, no now, the led is working, and i have no more the problem on seeing a bit of the other ledring move on this led ring.

But (Yep, there is always an anoying ''but''

to explain, the not working led was the last one of the led ring 9 (the number 144)

now, when i turn the encoder all the led are lighting, but for the ledring  11 12 13 14 15 16, all the 15 first led are lighning correcly, but the last one, is less uliminated, and so the precedently not working light is also lit, and for each one of the lasts led lightning, it is more lit (i an nore sure it is really clear)

[Lamouette]

i am using the uln2803

[latigid on]

It's not too clear, but it seems you now have a problem with the matrix not being properly connected and the rows are trying to sink through other channels. 

Are there any soldering issues around the 220R networks or the driver ICs? You could also try swapping the drivers to see if it's located on those chips.

[Lamouette]

i removed the ULN2803, i used direct linking, now : all the 16th led of the ledrings 11/12/13/14/15 are lighting in continue.

and if i move any encoder, i can see all the led lighting really low all around the led ring.

 

[latigid on]

Did you revert the sink mask (the driver chips invert the switching polarity)?

[Lamouette]

oh, no i forgot it, but anyway, all should have had the same state. i am redoing all the solder point of the ICs. i keep you updated ! Thanks !

[Lamouette]

and  same problem... i don't understand...

[Lamouette]

ok, if i understand correctly, and after continuity check, when the 16th led of the ledrings 11/12/13/14/15 are lighting they are creating a short on the led 16 on the ledring 9

[latigid on]

If you look at the schematic, the LEDs are arranged in a 16*16 matrix.

• Anode columns are driven by DOUT pulses on SR3+4, while cathode rows are sunk by SR1+2.
• All cathodes of each LED ring are connected to one DOUT pin (it's serving as a path to ground).
• Similarly, all anodes of ring(n) LED(x) are connected, so every ring(n) LED1 is connected, every ring(n) LED2 etc.

If there are any shorts to ground (e.g. if you scratch and bridge to the soldermask while installing an LED) then you can get unintended lighting. 

[Lamouette]

yes, i had understoood this, thanks for the explain.

and yes, i am pretty sure this is a scratch problem with the soldermask....

 

Evrytime the 16th of the faulting ledring is lit.i removed the coper all around, but it is still the same

 

if i turn the encoder until the 16led normally lit, all the other led have a really little brightness, once i pass the 15th , the 16th stop working and no other one is lighting

Edited May 8, 2016 by Lamouette

[Hawkeye]

The problem seems to be nicely isolated should be possible to find, don't give up! The fault is not necessarily directly around the LED pads, but could also be on any trace leading to that LED, e.g. a solderblob somewhere near the ULN or so? It should be possible to test with a multimeter, checking resistance to ground. There should be a difference in readout between the problematic LED and the other LEDs, when measuring the resistances of the "same" pads to ground (when the LEDs are not installed). Follow the corresponding trace... and good luck :-)

Many greets,
Peter