MB-6582 Questions

[cshesse]

Hi guys, I just almost finished building an mb-6582. I've got a couple of problems:

1) The sync LED doesn't turn on when I hit sync. Is it supposed to turn on? All the other LEDs seem to work, and when I hit the button, the LCD goes from the default screen to "Ens. Clock" so the button seems to work. EDIT: I tested the LED directly and it did turn on.

2) Some of the LEDs that are not supposed to be on are just barely on. I'm planning on reducing the brightness, so maybe this won't be a problem, but is that supposed to happen? I'm using some ultrabright LEDs with 1K resistors, planning on going to 3K because the lights are pretty bright. EDIT: I ended up with 68K resistors, which is nice and not bright, but also some of the LEDs don't seem to have even brightness, so I guess I'll have to keep it lower and brighter, unless there's some way to PWM the LEDs?

3) Well this isn't a question yet. My 4th pair of SIDs (from right to left) has a problem where the box doesn't turn on if I have the SIDs in there. Sounds like a short, but I haven't investigated where yet (I did notice 1V on the 5V/GND jumper, slowly increasing, none of the SIDs felt abnormally hot), and since I'm not sure if the socket somehow killed my SIDs, I want to be fairly certain before I put my only two spare SIDs in there. EDIT: I've now found this problem without the SIDs in, so it must be somewhere else on the board. I removed all the supporting chips (no SIDs or PICs) until the problem went away, then put them back, the problem did not return. I think it may be an intermittent issue with some of the supporting chips.

Thanks,

Chris

Edited December 25, 2009 by cshesse

[Wilba]

1. Sync LED isn't being used for anything.

2. This can happen when using ultrabright LEDs... it's an artifact of them being in a matrix. I noticed the same effect, it's most noticeable in the 8x8 mod matrix LEDs, where a single LED would have a "ghost" in an adjacent column and in the same row. My guess is that as each column gets scanned, there is still some lingering current sinking on the previous column, I don't know exactly why, it doesn't seem to affect the switches, i.e. there is never any false triggering of switches which are also connected to the same matrix. The only solution is to reduce the brightness by increasing the resistors, however, this can also lead to the other effect you notice, if you're running ultrabright LEDs with such low current, then this will highlight any variation in brightness... and also (I have to admit), cheap ultrabright LEDs from China are made for people pimping their cars and not really for control surfaces.

3. This sounds like your C64 PSU cannot deliver enough current on the 5V supply when using 8 SIDs, each SID draws 100mA on the 5V supply. I've posted solutions in other threads...

either replace the C64 PSU or reduce current on the 5V supply by switching the LCD backlight to use 9V supply. The best test to see what the problem is would be to measure current load. The easiest short-term fix is disconnect the LCD backlight (perhaps desolder the A/K wires on the LCD - they are nearly always pin 15/16) and then see what happens, you'll probably have it booting nicely with 8 SIDs and working fine, then you'll know the problem is purely the current load being too much for your C64 PSU, you can then try the alternate backlight supply idea, cut one track, add a wire, should be fairly easy... I recommend replacing the brightness trimpot with 50K for more control of actual current supplied to the backlight, and also learning how to measure current being supplied to the 5V rails AND also measure current supplied to the LCD backlight (after you do the voltage fix).

I will have to write up how to do this one day, a lot of people have this problem and it seems like I'm explaining a solution too often!

[cshesse]

Thanks for the help!

1. Nice to know

2. Well it's too late now, but do they have white LEDs of that size that cost less than 50c each and are NOT ultrabright? Because I would have bought them if I saw them anywhere.

3. The PSU claims 7.5VA for the 5V rail and 9VA for the 9VAC rail, so I guess it could still be a faulty power supply, it's very intermittent though, it usually works when all the SIDs are plugged in and I've not moved the box for awhile. Maybe it takes some time for the 5V circuit to warm up? (seems weird)

EDIT: it's a black C64 power supply, so I just took apart the 5V part and replaced the caps and resoldered the resistor and it seems to work now. Except that I didn't have a 4.7mF capacitor, so the leds are all wavy now.

Edited January 13, 2010 by cshesse