flickering LEDs and dimly lit LEDs after LED color change mod

[designlogik]

First off just so everyone I am pretty new to electronics with regards to troubleshooting and any deep understanding of circuitry.

Okay the issue I have ran into occurred after modifying a Roland MC-303. I decided I wanted to change the LEDs to be blue. After desoldering all the original LEDs and soldering in the new blue LEDs and powering up the device I found that there is now a noticable flickering of all of the LEDs which are on. This flickering can be seen in this youtube video https://youtu.be/k7DLOdJR9u0 The flickering in the video is worse than when viewing with your eyes but still noticeable. I am not sure what is causing this flickering, my guess is that it has to do with possibly the refresh rate of a signal being sent from one of the MCUs but I am not sure. It is a DC powered circuit so we can rule out AC fluctuations. I believe the flickering may have existed before the modification but is more noticeable with the brighter LEDs. Either way seems odd to me that this is happening, and any insight regarding why this is occurring would be great.

The second issue is that after the modification some of the LEDs stay dimly lit when they should be off. I have attached a photo of this. I've looked over the schematics which I have also attached photos of and the complete service manual can be found at https://archive.org/details/RolandMc-303GrooveboxServiceNotes The LEDs are on the panelboard. I could not figure out why this is occuring after tracing the circuit and reviewing datasheets for some of the ICs specifically the TD62785, 74HC245, TD62384AP, and 74HC138 I came up with nothing. It seems like there is some stray voltage on the circuit just slightly powering some of the LEDs on, not sure why changing the LEDs would cause this to happen or how to fix this.

Anyways any help or insight on these issues is appreciated.
If anyone needs anymore information let me know, I'll do my best to provide it

View attachment 65819View attachment 65820View attachment 65821

 

[tronitech]

Have you reviewed the datasheets for the LEDS? I know one LED is very much like another but things like Vf and If do vary.

Which blue led have you used?

 

[designlogik]

Have you reviewed the datasheets for the LEDS? I know one LED is very much like another but things like Vf and If do vary.

Which blue led have you used?

I've done some searching and was unable to locate a datasheet for the original LEDs. The LEDs I used for this modification were common 3mm blue colored diffuse LEDs with ~3.2 forward voltage and ~20mA forward current. The do differ form the originals, however I have successfully modified a different circuit which had very similar style LEDs and while there was some slight "voltage leak?" where a few of the LEDs were just very very slightly dim when they should be off, it was not close as bad as it is in this modification. I am still not sure why these LEDs are dimly lit, even on the other circuit. It doesn't appear to affect the core functionality of the device. Just wish I could pin point the underlying cause within the circuit and if it is not firmware related and instead has to do with part values I could change those parts.

I wonder if adding in some capacitors I could alleviate the flicker issue that I pointed out?

 

[tronitech]

I found a rather blurred and scanned copy of a datasheet here. The original SLR55VC LED's would appear to be 2.0 Vf and 10mA If.

Perhaps increasing the 22R resistors, R2 - R9, would help?

 

[designlogik]

I found a rather blurred and scanned copy of a datasheet here. The original SLR55VC LED's would appear to be 2.0 Vf and 10mA If.

Perhaps increasing the 22R resistors, R2 - R9, would help?

I actually had tried doing that before creating this thread, I tried using 100R in place of the 22R for R2-R9 but that did not make any difference. I may try to test maybe 560R or 1K just to see what the effect is, if any.

 

[designlogik]

I found a rather blurred and scanned copy of a datasheet here.

Okay I was searching using the part number located in the parts list of the service manual, didn't search the part number listed at the bottom of the schematic. The fact that they appear to be 2.0 Vf and 10mA If seems like that isn't a huge difference.

Edit: actually the SLR55VC are 5mm LEDs which are located under the 16 keys I changed those out for typical 5mm milky diffuse blue LEDs the LEDs I had searched for are listed as LNJ282RKEXE and I was unable to locate any datasheet for those.
On the other hand I did locate a datasheet for the SLR-325MC at https://pdf1.alldatasheet.com/datasheet-pdf/view/394765/ROHM/SLR-325MC.html and those are listed as 2-2.1 Vf and 20-25If. I also changed those to the blue 3mm

Seems odd that the LEDs I replaced have a lower Vf yet it appears as though there is stray voltage lighting the blue LEDs which have a higher Vf. My initial thought would be that the LEDs which are meant to be on would not be as bright and there would not be dimly lit LEDs which are supposed to be completely off. I don't understand.

 

[unclejed613]

you posted the main pcb schematic, but you showed a picture of what must be the panel board. could you post a schematic of the panel board? there seem to be no LEDa on the main board schematic. if we had a clue what was driving the LEDs, then we might better help you. it's possible the original LEDs were low efficiency (this device seems after all to use older TTL chips, so LEDs made back when they were more common would be less efficient, and wouldn't have enough "oomph" for leakage currents to cause any light to be emitted). there may also be clues in the schematic as to whether the original LEDs were "logic level" LEDs, which would have had an internal dropping resistor.

 

[designlogik]

you posted the main pcb schematic, but you showed a picture of what must be the panel board. could you post a schematic of the panel board? there seem to be no LEDs on the main board schematic.

The panel board schematic is attached and was attached when I first posted this thread. Is it not showing up?

 

[4pyros]

Almost sounds like one or two of the LEDs is backwords.

 

[designlogik]

Almost sounds like one or two of the LEDs is backwords.

The polarity of the LEDs is correct and was double checked during the modification.

 

[designlogik]

still haven't got around to changing out the resistors just as a test, need to order some resistors

 

[picbits]

Blue and white LEDs can be strange beasts. I've managed to dimly illuminate them by body resistance when inserting them into a "live" breadboard.

It could be that the leakage current when the LEDs are "off" is enough to dimly illuminate them. Possibly try putting a 1N4148 diode in series with one (and possibly a 2.2k resistor in parallel with the LED) and see if the extra voltage drop helps at all ?

Don't remove the voltage dropping resistor, just have the following

+V -> dropping resistor -> 1N4148 -> blue led with parallel resistor -> ground

 

[KJ6EAD]

Since the new LEDs require more voltage for a given current, not less than the originals, I'd be looking to reduce the current limit resistors R2-R9 to a value that yielded the same average current as the original LEDs were getting.

 

[unclejed613]

The panel board schematic is attached and was attached when I first posted this thread. Is it not showing up?

the schematic is labeled "MAIN BOARD" , and shows arrows from several connectors, whose destinations are labeled "PANEL BOARD".


edit: sorry, i found it now. there was a space between the panel picture and the main board schematic that looked like a blank spot....

 

[unclejed613]

i agree with picbits, see if adding a diode helps, and increase the dropping resistor values. more efficient LEDs means you will get light from leakage currents. you might also try checking any electrolytic bypass caps at the chips, you could be seeing switching spikes in other gates being coupled to the LED lines. the bypass caps reduce the amount of noise that can cause that.

 

[circuitspecialists]

Blue and White LEDs require a higher forward voltage (At least double the voltage ) than RED, GREEN, or YELLOW. This could account for the issue you are having.