Shorts between CS board and base board

[pingosimon]

On my MB-6582, some of the buttons are being weird (Most noticeably SID 3 and 4, and sync and CC), and a whole row of LEDs is not working (the bottom row of the matrix).

I tested all the connections between the baseboard and CS board and found the following shorts:

JD1 - D0 D1, D3 D4

JD2 - D0 D1 D2 D3

JD3 - D3 D4

JD4 - D2 D3 D4

JD8 - D6 D7

As far as I can see, there's no reason for them to be shorted. The soldering is fine where I'm testing. I tried to trace back a few of the lines, but I don't really know enough to fix the problem that way.

All the rotary encoders and the rest of the LEDs work.

[Wilba]

It is probably the LCD shorting against pads on the CS PCB.

Detach the LCD and if the problem magically goes away, reattach using some insulation between the LCD and the PCB, such as sticky tape.

If you have trouble finding sticky tape locally, I can post you some. :thumbsup:

[pingosimon]

Nah, that'd be too easy for this particular box. (It's the same one that's been giving me all sorts of problems). I disconnected the LCD and it's still acting weird.

SID 1 2 3 seem to work fine. SID 4 selects SID 4, but it also brings me to the 'Ens Clock" menu.

The Sync button selects SID 4, and opens the Ens Clock menu

CC selects SID 3 and opens the Ens Clock menu

Edit elects SID 2 and opens the Ens Clock menu

Play selects SID 1 and opens the Ens Clock menu

The bottom row of the LED matrix is still broken.

[Wilba]

'Sync' button is supposed to open the Ens Clock menu.

Also there's a symmetry to these effects....

Sync and SID 4 share the same DIN pin.

CC and SID 3 share the same DIN pin.

Edit and SID 2 share the same DIN pin.

Play and SID 1 share the same DIN pin.

These 8 switches are in two groups of four... each group sharing the same "column" of the switch matrix.

See the two top-layer tracks connecting this group of LEDs/switches to the 8x8 LED matrix to the left... those are the two "columns" of the LED matrix being used. Check for shorts there.

Combined with the fact that the bottom row of the 8x8 LED matrix doesn't work... this sort of suggest some short down there, between two pins of a LED. Note I've also heard of totally bizarre problems like this when a LED was shorted, or put in the wrong way.

So check soldering around that area... perhaps also test all the LEDs are working in isolation (supply 5V through 1K res. to the LED leads) and are also correctly oriented.

[pingosimon]

'

See the two top-layer tracks connecting this group of LEDs/switches to the 8x8 LED matrix to the left... those are the two "columns" of the LED matrix being used. Check for shorts there.

I think I've found the track for SID 1 2 3 4, but I can't find the one for the other four buttons. How exactly do I check for shorts there? There are no shorts between the two pins of any one LED.

I tested all the LEDs in question. One of the LEDs in the bottom row of the matrix is broken, so I'll replace it. Also, when I light up SID 1, the Play LED lights up, SID 2 and Edit both light up, SID 3 and CC both light up, and SID 4 and Sync both light up.

[Wilba]

If the two "columns" of the switch matrix are shorted, then that would also explain the LEDs in that top group of 8 also being wrong... as it's a combined switch/LED matrix... the LED above each switch in that group has common cathodes connected to the same "column" as the switch.

I'm suggesting you look near the bottom row of the 8x8 LED matrix for shorts since that row of LEDs is also not working, but that could also be just a problem with the output for that row. You may be dealing with more than one problem here.

Check that none of the switch/LED matrix "column" pins/tracks are shorted to each other. These are the far left outputs, what connects to the 8 transistors on the base PCB. If you find a short between these, it might not be in that area, it could be elsewhere on the PCB.

oops I forgot to read your first post more carefully:

JD1 - D0 D1, D3 D4

JD2 - D0 D1 D2 D3

JD3 - D3 D4

JD4 - D2 D3 D4

JD8 - D6 D7

shorts on JD1, JD2, JD3, JD4 <-- expected behaviour. These connect directly to the encoders, and depending on the position, sometimes one or both of the encoder pins will be shorted to ground, therefore you will sometimes see both pins shorted with each other, through ground. Nothing to worry about, especially if the encoders are all working.

JD8 - D6 D7 <-- that's definitely the problem with the SID 1-4, Sync CC, Edit, Play switches/LEDs.

Can you check that these are not both shorting to ground as well?

[pingosimon]

uhh...embarrassing...I found the little line of solder that was shorting JD8 D6-D7. It was under the ribbon cable. I fixed it, and that was definitely the problem. Thanks so much!

The bottom row of LEDs in the matrix is still broken, but I bet replacing that one broken LED will fix that.

EDIT: Nope...this is weird:

The LED itself wasn't broken. It's the bottom row, 3rd from the left. When the LED is soldered to the board, it doesn't work at all, even when I connect a 5v source directly to it. When I take the LED out of the board, it works just fine connected to the 5v source. I tried 3 different LEDs, all with the same result.

Edited October 29, 2010 by pingosimon

[Wilba]

Is there a short between the two pads where this LED won't work? That would explain why it doesn't work.

[pingosimon]

Nope, no short there.

[Wilba]

OK... hmmm... maybe there is a short between the anode pad (which connects to the "row") and ground.

Let's say you're testing using a 5V/Ground from the MB-6582... then putting 5V on the anode of the LED (when in the PCB) might be shorting to ground instead of going through the LED and to ground via the LED cathode.

But that won't explain why other LEDs in that row work in isolation.

Can you take out the 74HC595 from the base PCB and try testing the LEDs by putting 5V into the IC socket pins which connect to JD7 and JD7 (via the resistors and resistor/transistors). i.e. simulate logic high outputs of the 74HC595. You may need to consult the PDF to show you which pins relate to JD7/JD8 pins.

Also, which LED in the bottom row is not working?

Does the rest of the bottom row work (when the app is running) when you take out the non-working LED?

[pingosimon]

Let's say you're testing using a 5V/Ground from the MB-6582... then putting 5V on the anode of the LED (when in the PCB) might be shorting to ground instead of going through the LED and to ground via the LED cathode.

I'm getting the 5v from an optimized PSU that I built for another Midibox.

Can you take out the 74HC595 from the base PCB and try testing the LEDs by putting 5V into the IC socket pins which connect to JD7 and JD7 (via the resistors and resistor/transistors). i.e. simulate logic high outputs of the 74HC595. You may need to consult the PDF to show you which pins relate to JD7/JD8 pins.

You mean JD7/JD8, right? I'll try to figure that out when I get my next chunk of free time.

Also, which LED in the bottom row is not working?

The 3rd from the left.

Does the rest of the bottom row work (when the app is running) when you take out the non-working LED?

Nope, same result when there is no LED in place.

[pingosimon]

I tried the three bottom 74HC595 sockets and found this out:

(I'm using the 5V from the MB6582 baseboard now)

On U22 (The 74HC595 that goes to JD7) I connected pin 8 to ground, pin 15 to 5V makes the whole LED matrix blink, but the bottom row still doesn't work. Same result whether or not an LED is connected to that bottom row, 3rd column slot.

BUT

Just to make sure, I used the 5V rail (from the MB6582 baseboard's power LED this time) to test the bottom row LEDs. As soon as I turned on one LED with this method, the whole row worked! I plugged the 74HC595 and the power LED back in, and it still works!

I don't understand, but I'm glad to have a fully functional MB6582! Any idea what happened?

[pingosimon]

Nope, now it's broken.

It was working perfectly for a couple minutes. I left it on and walked away for a few seconds, and when I came back, all the LEDs are off, and the LCD's backlight is on and all the pixels are activated. That happens every time I power it on.

Should I make a new thread for this problem?

Edited October 31, 2010 by pingosimon

[Wilba]

So are you saying the LCD is now NOT working?

This could be related to a power supply problem.

Have you installed any SIDs yet?

Often if you have a slightly broken power supply, it can't supply enough power for 8 SIDs and everything else. If you have SIDs installed, remove them all and reboot, see if the LCD works again.

[pingosimon]

Oh hey, switching the power supply worked! That's weird...I wonder why a good power supply just broke all of a sudden.

Thanks a lot!