claudeheintz
commented
6 years ago I ran a test with 2 ESP32s one configured as Input_To_Network as an access point, the other as a receiver set to log into the first one in station mode. The access point took DMX input and broadcast it as Art-Net. I had to change the _wifi_config->input_address to 10.110.115.255 from 10.255.255.255. I'm guessing this is to match the subnet mask although I believe that the wider broadcast address should have worked as well. At first, I got some strange partial packets. But, then, partly to test the serial monitor, I added a printout of the number of slots and the level of a couple of addresses to the checkInput function in the main sketch. The network output settled down and worked correctly. I haven't had time to investigate. However, in loading both the receiver and transmitter, serial monitor functioned as expected. For this particular test, I was using a Microsoft Surface 4 with Windows 10 Pro. Initially, the serial monitor was set to 9600 baud and the sketch uses 115200 and that, of course, caused an issue with jibberish. But, once the baud settings matched, there was no other sign of a problem with the diagnostic serial. You don't say what computer you are using (and there shouldn't be a problem with 115200) but you can always slow down the baud in the sketch to 9600 and see if that makes a difference.
When I get a chance, I will re-visit the test using sACN. I recall an issue with multicast in access point mode. But, I don't remember specifics or for sure if it was ESP32 and not ESP8266.
Andreadj
tapi35
This is kind of 2 issues, but they might be related. I have 2 identical ESP-12N development boards each with it's own MAX485 driver. I am trying to set up one as a transmitter and the other as a receiver. I am using the ESP-DMX example code. I am not using the configuration utility, but instead changing the values in DMXwifiConfig::initConfig. I can get the boards to talk to each other, however I cannot check any debug information on the transmitter.
The only differences between the 2 in code is the following lines. Receiver:
_wifi_config->wifi_mode = STATION_MODE;_wifi_config->protocol_flags = MULTICAST_MODE | STATIC_MODE;Transmitter:_wifi_config->wifi_mode = AP_MODE;_wifi_config->protocol_flags = MULTICAST_MODE | INPUT_TO_NETWORK_MODE | SACN_MODE;I have uploaded the same code onto both boards to check if it was a hardware issue, but it makes no difference. If the board has the flag
INPUT_TO_NETWORK_MODEthe serial monitor no longer works, just outputs garbage. I also ruled outAP_MODE, the following outputs correctly:_wifi_config->wifi_mode = AP_MODE;_wifi_config->protocol_flags = MULTICAST_MODE | STATIC_MODE;But this does not work for transmitting over the network.Despite the serial monitor not working, the code appears to be functioning correctly (or at least partially). The transmitter does set up an access point with the correct parameters, and the receiver connects to it instantly. It seems to be dropping frames though.
I am using qlcplus and a DIY OpenDMX USB device that is transmitting at 30 frames/sec. The ESP transmitter LED is blinking steadily when data is present. The receiver LED is completely sporadic, and the DMX data coming out of it is as well. Both units are running at 160 MHZ, and are located ~10cm apart. They are running off the same 2 amp 5v power supply. I have tested using both
SACN_MODEandARTNET_MODEand neither seem to work better. I have not tried unicast yet, and would prefer to stick with multi if possible.Has a similar setup worked successfully for anyone else? Wondering if there's something that i am overlooking?
EDIT: Suspecting the dropped frames were an issue with the
AP_MODE, I set up a consumer router as an access point and ran both the transmitter and receiver inSTATION_MODE. The transmission became much more stable, and is actually usable now. I suspect that there is some overhead in thesoftAPthat is unable to handle DMX.I still cannot use the serial monitor, but the system is functioning.