Vixen to FPP fseq issue, renard data corrupted.

[Rocketman1]

I have searched and searched for a solution, but no luck

I have a simple on/off sequence created on Vixen, and exported it to run on FPP, and it appears to work fine.
If I replace the on/off functions with a ramp, twinkle, ect. the data seems to be corrupted going to the Renard controllers.
The more complex the sequence, the worse the corruption seems to be.
If I move the RS485 from the Pi to the computer, the sequence runs correctly.
Using the test function in FPP, everything works as expected.
I have killed all software on the PC, so there should not be any interference on the wifi into the Pi (seems others have been bit by this) - I don't have any inputs enabled anyway on FPP.

Details:
Vixen 3.5u4
Pi B+ running latest version of FPP, using recommended 2.5A power adapter.
RS485 dongle (DSD TECH SH-U10 USB to RS485 Converter with CP2102 Chip)
connected directly to Pi, going to two Renard TR8 controllers in series (16 channels).
Cat 5 cable is used, ~3' Pi to first controller, and 10' between them to the second one.
I have tried both termination and unterminated at the Renards, no difference.
The yellow framing light never blinks.

I am currently waiting on a powered hub, thinking there could be insufficient voltage at the USB port.
Also planning on building a similar test sequence on xlights, but not had the time to do that yet.

Any other ideas?

[mrGrumpy]

Suggest you add current to your Pi supply voltage. 2.5A is a little low for the B+ and unless you need the powered hub for something else, I think it can be omitted, and upsize your Pi supply to may be a 5A supply.

Ramp/Twinkle issue . . . could be your dongle, but also the Export Timing. If you are using 25ms, try 50ms. I've not found a case where 25ms is needed, so until you actually need 25ms timing, use 50ms.

Good luck, and hope the simple timing change works for you.

[MikeKrebs]

Another thought...The dimming (and therefore the ramp) on the TR8 relies on the zero cross. Is it installed properly and working?

Can you try a really long sequence of the same effect? Make sure your dimming curves are 100 percent and then set the ramp for like 30 seconds. What does that look like?

Also, did you set the port speed to be the same as your controllers are flashed and are the controllers both flashed with the same firmware?

[Rocketman1]

Mike,
I don't think the zero cross is the issue, two reasons: 1. the zero cross indicator is on, indicating the PIC is getting the signal, and 2. the test sequence works fine when I move the USB-RS485 dongle from the Pi to the PC, and run the test from Vixen.
Grumpy,
I have been on business travel the past couple of days so have not had opportunity to do any other testing. I think I have Vixen set for 50mS, but will confirm when I get home tonight, and I received a powered hub that I am going to try to see if the 5v is too weak coming out of the Pi for the dongle and see where that goes. Next option is to order a different dongle with a allegedly better chip in it. If that works (or if it doesn't at all), I will try to locate a higher output adapter to power the Pi.

Thanks for the thoughts.

[MartinMueller2003]

There is a MAJOR difference between the PC and the PI. Windows vs Linux. Unless the proper drivers are installed on the PI, it is possible that your device will not work properly. You did mention you are using the 2102 chip which is supposed to work on the PI. You also mentioned that using the FPP test screen you were getting the expected output. This does mean that there may be a problem in the FPP code. Do you have older sequences that play correctly?

[Rocketman1]

I don't have any other sequences than the ones I have created this year (new at this). I generated a sequence in Vixen, saved it, and exported it to .fseq Then uploaded the file to the Pi.
Checked the Vixen Tools/Options and it says update interval is 50.
Tried using a powered USB hub, no difference...
Going to start working up a similar xLights sequence and see how it behaves.


Just to be clear, this is a snippet of what the test sequence looks like in Vixen sequencer, and the output response on the Pi has brief flashes from lights that are not supposed to be on:
Annotation 2019-10-09 191813.png

[mrGrumpy]

I believe the Tools/Options is the computer timing, not the .fseq timing.
If you run correctly off the computer and the Falcon TEST seems to work correctly, then I suspect if is how the file was exported. Try exporting that sequence again, but in the Export phase, name it ??????V2 or something, and use the basic settings:

Format: Falcon Player Sequence
Timing: 50ms

Do not unclick any Controller
Leave Generate Controller File unclicked.
= START =

Load this new export and test.

[Rocketman1]

ARRRRHHHH!

So I have tried using xLights to generate a .fseq file, same glitchy result with lights flashing on when they are not supposed to. One nice discovery was xLights generates the .fseq file at the same time you save the sequence - didn't know that.

Got a new RS485 dongle, this one with a FTDI chipset. Still no change.

I did a clean install of vixen as there seemed to be some things missing, like import, and the 'generate controller file' wasn't in the export box. Performed an export per your guidance above, and tried many other export formats like falcon player sequence 2.6+, 25mS, 50mS, 100mS. Nothing seems to make any difference, I still get the output errors...

I am so aggravated I am at the point of giving up on the Pi and just running it from the laptop.

[MartinMueller2003]

I suggest asking on the Falcon forum

[Rocketman1]

Thanks, heading there next...