vespine v1.0: Oscilloscope fun. PWM in action.

I've made some progress tonight after a lot of poking around with the CRO I borrowed. Turns out, I've probably wasted quite a lot of time investigating my LED "issues", what I've worked out is the TLC5940 boards I've made are not actually working how I thought they were. It turns out, the signals to the 6 chips on each board is degrading to the point where the chips are not functioning properly, I probably should have picked up on this earlier, but i was so convinced it was the LEDs that I neglected to check the chip boards until now.. I mean, they were "working" enough to convince me it was ok, but not enough to actually light the cube up properly. It's only once I stared programming the chips to light separate LEDs that I realised the problem.

At this stage, I'm not sure, I'll have to sleep on it, quite possibly I will have to re do both tlc5940 controller boards, make them better, use better wire for the connections perhaps..

You can't really make it out very well but in the below, ten LEDs are supposed to be lit and the rest not, the ones which are lit but dim are actually flickering, which you obviously can't see in the photo.

This is the traces of 2 of the 8 layer select MOSFETS doing their thing. I was happy to see this, it looks pretty good. You can see the high segment on the bottom trace followed by the high segment on the top trace, this is where the mosfets are "ON". The remaining 6 mosfets would fill the rest of the gaps where the small bumps are, but the CRO only has 2 probes.

This is interesting, i took measurements of what comes out of the TLC5940 output at various intensity levels. This 1st photo shows 2 traces, the bottom is at full brightness, the tlc5940 has 12 bit brightness so that equals to 4095. The top trace is at 3000, so you can see roughly a quarter is high and the remainder is low. Of course the tlc is a current sink so low is actually ON.