LEDs remain on (x0xi0 build) & sequencer problems

[effegee]

I'm building BC Audio's x0xi0 with custom PCBs & custom enclosure (the complete kit).

I've tested VCO1 & 2, VCF, VCO sync, Xmod, etc., on a 'scope so I know the analog circuit is good. I'm having problems with the digital section.

I'm at the step of testing the VCA by programming and playing a pattern. I was able to get some (sporadic) function of Pattern Edit mode, enough to figure out where the Mode encoder is pointing. I can't program a complete pattern, etc., but I first want to figure out why I have Step and sequencer LEDs lighting up when they shouldn't be. When I power it up, several of the LEDS light up and remain on (usually 75% of the left side LEDs). This happens whether the microcontroller is in its socket or not. I was told that the step sequencer LEDs should not be lit when powered on without the microcontroller in place. I want to solve the LED issue before I sort out why the sequencer is not working properly. I don't know if the two issues are connected.

I've tested voltages and they're good. All of the sequencer ICs are placed in proper orientation and are getting 5V. I've scanned the PCB both sides with a magnifying glass. I've tested continuity of LED_latch, SPI_CLK, SPI_IN and they're good. (I don't know how to test for continuity of SW_LATCH.)

At this point, when I turn it on without the microcontroller, 75% of the LEDs and tempo LED remain on until I touch pin 32 of the microcontroller socket, then they all go off. If I sweep my finger across pins 1-20, then all of the LEDs turn on, sweep again and they turn off.

Can anyone help me?

[guest]

it is normal for the leds to come on randomly
when the there is no micrcontroller

[effegee]

Thanks for the reply. It's good to know it's normal behavior. Unfortunately, it does the same thing when the micro is in place.
I put the micro back into the socket and when I power it up, the random LEDs light up solid, including the tempo LED. Then I touch pins 1-8 of the micro (I assume, grounding SPI_IN and SPI_CLK?) and the LEDs turn off and the Tempo LED starts flashing and Step 1 LED lights up solid (in Pattern Edit mode). The next button works and I can write a pattern sometimes, but none of the black key LEDs light up when I press the corresponding switches. I'll have to read up on what's normal behavior in Pattern Edit mode, but for now I still have the random solid LEDs on power up.
Any ideas as to why?

[guest]

does random mode work
does keyboard mode work

[effegee]

I don't know if Random or Keyboard mode work because I don't have the little mixer board PCB hooked up yet to listen (it's a x0xi0 thing). I was, however, able to program and play a pattern to test the VCA. Unfortunately, I'm not able to detect a pulse from D36(+) on my vintage 60's 'scope. (I thought I might not be using it correctly, but I can see the trigger pulse from my TR-606 and Gate from my TB-303 on this ghetto, garage sale scope, so wouldn't the x0x's pulse also show?) I also probed pin 3 of VR8 (volume), but I see nothing on the scope and hear nothing when probing with the tip of some earbuds, so something is not connected here.

I still have to touch the micro pins to shut off the LEDs so that the digital section will work.

[guest]

if the gate out is going to 12v and back down
then the microcontroller must be partially functioning

but touching the micro pins
shouldnt effect the leds
so im guessing that there is a loose connection
in the microcontroller socket
try rehitting some of the solder joints
either that or some of the microcontroller pins could be fried

[effegee]

I didn't say that the gate was going to 12v. I see the note and step LEDs flashing when the pattern plays, but I cannot detect any pulses at D36(+). Wouldn't that mean I'm not getting a gate signal? I'm just trying to understand here.

[effegee]

I just discovered that I had D33 in backwards, so I just reversed it, but nothing's changed. Do you think I fried something?

[effegee]

Just after power up and the random LEDs are on solid, I probe pin 9 and get 4.69v. After I brush the pins with my finger, the LEDs go off and the sequencer Pattern Edit is ready, pin 9 goes to 5v. Hmm...

[effegee]

Ok, now I see that the 7805 (IC20) is not putting out 5V when I power up. Probing pin 3, I get around 4.68v until I do the finger sweep.

[effegee]

With the micro out of the socket, no random lights now and pin 9 of the micro is at 5v. I reflowed all 40 pins of the socket and inserted the chip, but it's still back at 4.68v with the random LEDs lit. Do I have a bad micro controller?

[effegee]

Another update:
I seem to be getting a more consistent power up without out the random LEDs and it goes into Pattern Edit mode as intended. I've discovered that if I switch it off and then back on again, the LEDs come on again and it's locked up. However, if I turn the unit off, count to 5, and turn it on, it behaves normally. Counting to 3 or even 4 is too soon. The magic number is 5.

Is this normal behavior?

(Now to figure out why the VCA is not working...)

[guest]

something is wrong with your leds or led resistors
or maybe the led latches

double check that all the led resistors are the right value
they are drawing too much power
and causing the microcontroller to reset itself

[effegee]

I asked Brian for green LEDs instead of the stock red ones. I also asked for red, orange, yellow and white ones for the 16 steps (a la TR-808). He sent those and 220 ohm resistors. For the 4 whites I ended up using 10k resistors to match the brightness of the other LEDs. He told me that was fine.

[guest]

are your led latches socketed by any chance
you could take them out and see which leds are doing it
or it might just be when a bunch turn on

what is the input voltage to the 7805
when the device is failing
and what is it when its ok

a powersupply with too little power could also cause this