Bit Angle Modulation (BAM) in a PIC

[Mike, K8LH]

What "thing" are you talking about tossing Darrel?

I wrote and simulated the ISR driver for a 5-pin 20-LED demo' but that's as far as I got. You want to see that part Sir?

<added>

Tell me what chip you want it for and I'll give you a simple 5-pin (rb0..rb4) 20-LED sample Saturday that you can breadboard and play with.

Mike

[Darrel Taylor]

What "thing" are you talking about tossing Darrel?

At this point, I'd like to toss Charlie Allen himself.

I wrote and simulated the ISR driver for a 5-pin 20-LED demo' ...

Is that with MIBAM? I'd love too.
For whatever chip, in whatever language.

Even if it's not MIBAM, it might give me an idea that could save Charlie's life.

Thanks again Master,

[Mike, K8LH]

Yes, it's a sample ISR engine for MIBAM within an assembly language 'skeleton' program for 16F886 with 20 Charlieplexed LEDs on 5 pins (RB0..RB4). Just needs config lines, an init section, and short main section to manipulate the led[] array. I was interested in seeing if MIBAM could be implemented on a Charlieplexed matrix and after I wrote and simulated the MIBAM ISR 'engine' I got distracted by other projects.

I will send you the Charlie-MIBAM skeleton in a PM and work on those missing pieces so that you can actually try something on some hardware, if you like.

Mike

[Mike, K8LH]

Someone asked about the diagonal line on my Charlieplexed display sketch. It's what I refer to as the "float" pin and it's connected to the LEDs that occupy positions in the matrix that would otherwise share a common column and row (RB0/RB0, RB1/RB1, etc.).

I apologize if the diagonal line has thrown you off. It's a remnant from some of my sketches of Charlieplexed 7-segment displays.

Both drawings below are electrically the same but the second drawing below may make more sense. Basically I just slid those LEDs in the bottom row up into the empty spaces in the matrix.

Kind regards, Mike

[Mike, K8LH]

Darrel,

I was just wondering -- is MIBAM even necessary on a multiplexed display due to the long pauses between LED activation. You know, column one LEDs would be BAM'ed then turned off while we sequencially BAM column two LEDs, then BAM column three LEDs, etc.?

[Darrel Taylor]

Hi Mike,

I wholeheartedly agree.
MIBAM with multiplexed display doesn't seem like it's worth it, because you loose so much of the available dutycycle.

Thanks for the explanation of the diagonal line.
I did get your example working, with only 8 leds (2 columns).
I made enough changes to not know if it's actually the way yours works or not.

But I don't see a way to incorporate it into the MIBAM routines (because of the duty cycle losses), so I'm officially tossing Charlie out the window.
Good thing I live at ground level.

I'll complete the modifications for your previous idea (full port updates) and get this thing done.

Thanks for all your assistance Mike, much appreciated.

[Mike, K8LH]

Yes, LED duty cycle is decreased with any multiplexed display and while we normally pump up the "peak" current to get better "average" current (and brightness) we can't do that with a Charlieplexed display (or any other multiplexed display that uses "direct drive" for matrix columns or rows).

Anyway, I thought you might enjoy taking a peek at a multiplexed 8-channel RGB display demo' (24 discrete LEDs) with 256-level BAM driver and a pretty amazing 868-Hz refresh rate. The display is fashioned as a 3x8 matrix (33% duty cycle) so biasing the LED "peak" current to approximately 3x desired "average" current will be necessary. This assembly language program just contains the driver. No animation or anything like that, yet.

Regards, Mike