Replies: 13 comments 27 replies
-
|
Hi
Are you sure about it? Because then CE0 and CLK are swapped. Compare theses picture (color of cables are the same). |
Beta Was this translation helpful? Give feedback.
-
|
Sorry, I didn't list them in order, but they are connected as such: ESP32 -> RPi5 |
Beta Was this translation helpful? Give feedback.
-
Check again RPi5 diagram because I think it should be GPIO 8 ...I use only my old photos and the same connector when I need to connect it again so I don't remember exactly. |
Beta Was this translation helpful? Give feedback.
-
|
This is the pinout I was referencing:
Should I be using CE1 (GPIO8)?
|
Beta Was this translation helpful? Give feedback.
-
|
I think this diagram is wrong |
Beta Was this translation helpful? Give feedback.
-
|
Maybe I am looking at it wrong, but the close up pictures appear to be using the same pins that I am using? |
Beta Was this translation helpful? Give feedback.
-
|
CE0 is GPIO8 not GPIO7 as the diagram states. For the rest as I mentioned you can compare your connection to mine and also you can post pictures of your setup. Also I dont think default SPI speed is 20000000 but 25000000 in HyperHDR, not sure it is related. But enabling "invert" in the HyperHDR SPI driver properties could be related: disable it. |
Beta Was this translation helpful? Give feedback.
-
|
I messed with the speed and inverting the signal as desperate hopes to get any kind of output. I can switch them back and provide the logs if that would help? Here's my current wiring setup:
|
Beta Was this translation helpful? Give feedback.
-
|
The connection looks OK. HyperSPI wont work with "invert" option, but non-standard speed should not cause problems for esp32-s2. Try: 25 000 000, 10 000 000 and 1 000 000. |
Beta Was this translation helpful? Give feedback.
-
|
BTW, are you using custom firmware with multi-segment enabled? justinj57@75a00ac But in HyperHDR you have an different (and odd) number of LEDs configured, which can lead to unpredictable behavior (worst case scenario: communication works, but esp32-s2 crashes after receiving the first one from the out-of-scenario frame and restarts, therefore the statistics are empty). DSECOND_SEGMENT_START_INDEX is not something that can be set as you like. For testing use the regular released firmware. |
Beta Was this translation helpful? Give feedback.
-
|
There may be a GPIO fault, as the pins in RPi5 are still not 5V tolerant, although some can survive 5V if special conditions are met. If this is the case, it is difficult to recommend another board for HyperSPI driver because it will not change anything. In the "worst" case, there is still HyperSerialESP32, which in the case of esp32-s2 is not bad because its CDC is very fast and at least you can test your setup. RPi5 also has a second SPI interface available, but I haven't tested it yet |
Beta Was this translation helpful? Give feedback.
-
|
To be sure please send me your exported HyperHDR configuration (general tab) that you are using. I will test it on my RPi5. Do you have RPi5 4GB or 8GB and fully updated OS (Bookworm)? |
Beta Was this translation helpful? Give feedback.
-
|
Not sure if it can be related raspberrypi/linux#5696 but it should be fixed now |
Beta Was this translation helpful? Give feedback.
-
|
Here's my configuration. I have the second SPI interface active here, currently trying to see if I can get that to work. I have a 4GB RPi 5 |
Beta Was this translation helpful? Give feedback.
-
|
/dev/spidev0.1 is still first interface, only Slave Select is changed to other pin GPIO7 (lest just forget about that wrong diagram posted here earlier). If the first interface is damaged then changing just the slave device wont help almost for sure (unless only GPIO8 is damaged, very unlikely, if it is, it is damaged along with the rest). |
Beta Was this translation helpful? Give feedback.
-
|
OK, I restored your configuration. |
Beta Was this translation helpful? Give feedback.
-
|
Interesting. If you still have this hooked up can you send a picture of the connection? I'm starting to lean towards this being an issue with the RPi |
Beta Was this translation helpful? Give feedback.
-
I use the same connection from the beginning What OS and kernel version are you using? How did you install it why are you running it on root? |
Beta Was this translation helpful? Give feedback.
-
|
It's hard for me to tell what could be wrong, did you change the speed in the tester to 25000000? You can also check the eeprom version in RPi: vcgencmd bootloader_version. In the meantime, I tested RPi5 with esp32-s2 using even a 20 cm bundle of ordinary unshielded duponts and it also worked at full speed with less than 500 LEDs. I don't know if it has any significance, for me the ground goes together with the other cables, not separately. When it comes to setup, we use a different fan and USB drive vs SD card, although I don't think it will affect EMI. |
Beta Was this translation helpful? Give feedback.
-
|
Here's the bootloader version: Loopback test worked at 25000000. I'm losing my mind - Any chance you'd be willing to test this hardware if I mailed it to you and paid return postage? |
Beta Was this translation helpful? Give feedback.
-
|
If you are sure that HyperHDR configuration is correct this time that leaves a cable adapter e.g. crosstalk problem or esp32s2 issue which pinout required soldering. Maybe the time has come to consider flashing HyperSerialESP32 firmware |
Beta Was this translation helpful? Give feedback.
-
|
I wasn't ready to give up yet... I purchased some new DuPont connectors, since that was the only thing I hadn't replaced yet (made multiple cables, but using the same batch of connectors). Even though every single cable tested good for continuity, there must have been something screwy/intermittent with the connection to the pins on the Pi or the ESP32, because it now works! HyperHDR frames: 201 (FPS), receiv.: 201, good: 201, incompl.: 0, mem1: 0, mem2: 0, heap: 192736 Can't tell you how frustrating it is that this was apparently the issue. Extremely grateful for the help you provided and the work you're doing for the community. |
Beta Was this translation helpful? Give feedback.
-
|
I'm glad you didn't give up and managed to run the setup, congratulations! It was an unusual and complicated case. Now if we know that the problem was in the cable connection and at the same time there was continuity, I am almost certain that it was probably caused by unusual ground cabling or crosstalk because of routing of the clock and data signals locked in a separate isolated group without the ground alongside. So far, we could only observe problems when ground was not connected via GPIO: either a complete lack of communication or a very high number of transmission errors. And this despite the fact that the ground of two boards was still connected via USB. |
Beta Was this translation helpful? Give feedback.
-
|
I'll check the eeprom tonight. I did not try other speeds with the loopback but I'll do that as well. Hopefully I can get to the bottom of this |
Beta Was this translation helpful? Give feedback.
-
First off - thanks for all the work you've put into HyperHDR and all the supporting projects. Fully acknowledge that I'm likely doing something wrong here, but I am pulling my hair out and would be extremely thankful with any help I could get....
Hardware/Software Config
HyperHDR v19
HyperSPI
MS2130 USB Capture Card
3.3V to 5V level shifter
Raspberry Pi 5 (SPI Enabled)
ESP32 S2 Lolin Mini
SK6812 RGBCW
60A 5V Mean Well PSU
I have pins 7, 11, and 12 connected to the corresponding CE0, MOSI, and CLK pins + ground on the Raspberry Pi 5, with ~ 10cm cable. The ESP32 is powered by USB connection with the RPi5. I have double and triple checked all connections, made multiple harnesses, etc. I have also tried multiple ESP32 S2 boards.
I have not been able to get output on GPIO 2. When I try to run the performance check as outlined in the HyperSPI readme, I get 0 FPS.
HyperHDR frames: 0 (FPS), receiv.: 0, good: 0, incompl.: 0, mem1: 0, mem2: 0, heap: 193000
RGBW => Gain: 255/255, red: 160, green: 160, blue: 160
I think this means there's no communicated between the RPi and ESP32, but I am out of ideas on what to check. Here are the logs from start up to +3 min post start up. Please let me know if there's other info I could provide that may be helpful in troubleshooting.
Log: https://pastebin.com/Ksx81edA
Beta Was this translation helpful? Give feedback.
All reactions