I have successfully implemented a reasonably stable trigger circuit for the LightCrafter. I am using a 3.3v Ardunio Pro Mini to drive the trigger.
We load a sequence of 24 1 bit patterns of which the final frame is black (all off) and the first and 23rd are white (all on). We then send a quick burst of 23 trigger pulses (250-1,000 us interval) followed by a longer delay after the 24th pulse (16,566 us) which to our understanding would result in the first 23 frames being rapidly displayed followed by a black frame for a longer duration.
In practice, we are encountering considerable difficulties in terms of synchronizing the first frame with the first pulse, and we are encountering a shift of patterns in our sequence which would correspond to sporadic skip. We have tried 23 and 25 pulse sequences just in case we were encountering an "off by one" error, with no good outcome.
Which leads to one comment and several questions:
I feel that it is essential to add to the documentation that pin one (1) of the trigger connector is a voltage input to the LightCrafter. Until we inferred this from the schematics it was essentially impossible to use the trigger until this was resolved. Secondly, driving the input trigger does require a bit of current and it is best to put a 3.3v gate or MCP pin to drive it.
In terms of questions:
>> Is there a preferred pulse duration for the trigger? We are currently using 10 microseconds.
Any pulse width of more than 100 nano-second should work.
>> We are presently using the GUI to setup and run the LightCrafter. In external trigger mode, does the start button arm the trigger or start the sequence.
The LightCrafter internally configures itself for pattern sequence when the start command is issued.
>> Regardless of the trigger rate (250-1,000 microseconds) the LightCrafter seems to "skip a beat" every second or so, since this is very sporadic it is very difficult to pin down. Is this a known issue?
Please make sure that the input trigger frequency does not exceed maximum allowed frequency for the current configuration.
To know the max frequency/min period, do a "Get" pattern sequence setting and look for "Trigger Period" value
>> Is there any way to determine when the pattern is at the initial frame? Due to "skipping" our patterns shift through the timing sequence, throwing off our measurements and timing.
>> Is there any means of re-initializing the sequence so that we can reliably match a trigger pulse (in or out)?
We have found a bug in the LightCrafter FPGA and DM365 software that sometimes causes an extra trigger at the beginning of a sequence and/or a missed trigger in the middle of the sequence. We have isolated the problem and currently testing the fix. The fix involves new versions of the FPGA firmware and DM365 software. We will be releasing these shortly after our testing is completed.
Hi Pedro,
Are the new updates to the FPGA and DM365 available? If so where can we find them?
Daniel
Hi Daniel,
The software updates should be available by end of June. If it is available sooner, I will update in the E2E forum.
Best Regards,
Divya
www.ti.com/mems
Hello,
Please let me confirm whether the latest FPGA(V2.4.39) and DM365(V3.31) software have the solution for the above extra trigger problem.
Also, regarding the external trigger , would you please provide the information of the rising/falling time LightCrafter expect?
Best regards,
Rossinag
Rossinag,
Yes. The trigger issue is fixed in the latest FW versions.
We do not have guaranteed rise/fall times for the trigger signals. However, as Sureshkumar said in March - any pulse width longer than 100 ns should work. I would say that you should make sure that the rise and fall signals are monotonic, and that you probably should avoid extremely slow rise and fall signals.
Pascal