Boards are looking good gizmo! Let me know if you have any issues sourcing the headers.
I just merged the initial port to the asynchronous web server into the master branch. LOTS has changed, so please cruise the readme. You can't just simply flash the sketch. This WILL wipe your configuration, as I moved the configuration out of EEPROM. The Gitter chat room has been pretty active and I recommend jumping in if you need help with anything. Here's the current changelog for 1.5 beta:
### 1.5 (in progress)
- Migrated to ESPAsyncWebserver for the web server.
- Migrated all web pages to SPIFFS.
- Migrated configuration structure to SPIFFS as a JSON file.
- Added Gulp script to assist with preparing web pages.
- Changed default serial port for Renard / DMX to Serial1.
- Changed output type to compile time option. Check top of main sketch.
- Added mDNS / DNS-SD responder.
Just pushed a change that makes ws2811 output asynchronous. It's looking good on my 720 pixel matrix, but please let me know if you notice any issues.
Have others had issues with lag or choppingnesd using the ESP6823 controllers? How have you avoided lag issues? I built and tested 3 of the RenardPlus ESPwifiPixel boards and they had about a two second lag from what xlights was playing and seemed far less smooth doing fades compared to my wired controllers.
Sent from my iPhone using Tapatalk
grapas2 wrote:Have others had issues with lag or choppingnesd using the ESP6823 controllers? How have you avoided lag issues? I built and tested 3 of the RenardPlus ESPwifiPixel boards and they had about a two second lag from what xlights was playing and seemed far less smooth doing fades compared to my wired controllers.Sent from my iPhone using Tapatalk
Are you getting a lot of sequence errors? Is your xlights computer wired or wireless?
Sent from my XT1094 using Tapatalk
Sequence is delayed about two seconds. For example I hit play on xlights, and the lights come on two seconds later and then a stop xlights and the lights stop playing two seconds later. The errors are limited to short freezes. I do believe part (maybe most) of the problem is that I the xlights computer was wirelessly connected to the network. I was also sending a about 20k multicast channels out of xlights. I ran out of time to test if those two things fixed my problems. But I thought I would ask to see what everyone else is doing to eliminate lag. Would a wired connection to my wireless router and unicast likely fix things? I plan to have between 50 and 100 of these wireless receivers for this particular project.
Sent from my iPhone using Tapatalk
Also, what does the asynchronous web server provide?
Sent from my iPhone using Tapatalk
grapas2 wrote:... xlights computer was wirelessly connected to the network. I was also sending a about 20k multicast channels out of xlights.... I plan to have between 50 and 100 of these wireless receivers for this particular project.
That would definitely be your problem. You e1.31 source (xlights) needs to hardwired. 802.11 is simplex so all that traffic you're sending has to sit on your AP and wait til it can re-transmit to your controllers. I would recommend unicast as well. As for 50 to 100, that's not going to happen unless you have a very large wireless infrastructure with many access points, and more than likely you'll run into spectrum issues. I ran 15 off one AP last year, but haven't tested beyond that. How many channels will the nodes be handling? I'd try to use something like a nRF24L01 based solution if you really need that many receivers, provided you're not trying to push an insane channel count. 802.11 brings a lot of overhead with it, link layer handshaking, etc... whereas the nRF you can just fire and forget broadcast, but at a much lower rate (2mbps max per TX).
grapas2 wrote:Also, what does the asynchronous web server provide?
Sent from my iPhone using Tapatalk
The standard web server runs in the main loop, hogging all the processor while handling sessions. The async one handles traffic off interrupts, can handle multiple sessions, and doesn't block the processor in the main loop. Before, you had to terminate your e1.31 traffic before working with the web interface. Depending on your traffic, it would either crawl or not work at all. Now, you can interact with the web interface while handling a 720 pixel stream incoming just as smoothly as if nothing was coming in.
Thank you for the advice. I will only be pushing around 24 channels to each receiver. I have been working with Komby to see if I can get the RF1 to do the trick. It is having trouble with the battery system I am using but he has some ideas that may help.
Sent from my iPhone using Tapatalk
Latest commit removes the PPU options for pixels and will now utilize all 512 channels if you roll over into another universe. Previously, it was capped at 510 per universe when working with pixels. Also, Renard output now supports multiple universes and up to 480kbps baud rate (I can go higher, let me know if you need it). Check the Changelog for other goodies and notes.
Lastly, the group buy for V2 boards will be starting up in early August!
YAY! Thank You.
