Sharing ESP32 newbie's try

[intaker]

First of all, I really appreciate your the time and efforts put on Ignitron. By having a high school boy with Spark Mini amp who was 3D printing a foot lever to turn the amp mode dial, I've decided to give a go and this post is to share the experience and progress.

I do have experience with Arduino but it was first time with ESP32 and it was a good experience overall. A couple of late nights are good, isn't it?

Purchased some parts from Aliexpress. Mostly same but chose a bit bigger OLED as both of us need a special treat to eyes.

• ESP32

  • Used by Ignitron
    • https://www.berrybase.de/en/Pixelpdfdata/Articlepdf/id/136/onumber/NMCU-ESP32
    • Chipset : ESP-WROOM-32
    • USB-Serial : CP2102
  • Purchased
    • https://www.aliexpress.com/item/1005006974682840.html?spm=a2g0o.order_list.order_list_main.16.4b0d1802t3PS8t
    • Chipset : ESP-WROOM-32D and ESP-WROOM-32U(just for curiosity on the difference)
    • USB-Serial : CP2102
    • Comments from the seller
      □ Some of the chips have a bios that does not allow the flash mode 'QIO'. Changing it to 'QOUT' fixed it for all of them.

• OLED

  • Used by Ignitron
    • https://www.amazon.de/AZDelivery-Display-Arduino-Raspberry-gratis/dp/B074N9VLZX
    • Chipset : SSD1306
    • Screen : 128x64
    • Size : 0.96"
    • Interface : I2C
  • Purchased
    • https://www.aliexpress.com/item/1005005967766159.html?spm=a2g0o.order_list.order_list_main.10.4b0d1802t3PS8t
    • Chipset : SH1106
    • Screen : 128x64
    • Size : 1.3"
    • Interface : I2C
    • Chose bigger screen

• Foot Switches purchased

  • https://www.aliexpress.com/item/1005006163316744.html?spm=a2g0o.order_list.order_list_main.5.375b1802cbikM8
  • Bought 10 to be ready for defective parts and future replacement

It was first time to use VSCode + PlatformIO so I should say there was a short learning curve about a day(I do IT so pretty comfortable handling IDEs)

When purchased the OLED, did a basic research on display driver and replacing adafruit/Adafruit SSD1306@2.5.7 library with adafruit/Adafruit SH110X@^2.1.10 did the trick with the minimum code change.

Basic compile done, uploaded the compiled firmware and I could see the Ignitron logo on the OLED. Great!
It connects to Spark Mini in APP mode nicely. All seems good but always there is a twist.

Challenge 1
By being an Android user, tried to switch BLE to SRL using buttons but it just crashes and restarts. Even changed the code default to SRL but same crash and restart on Spark App connection. Enabled the debug and started to read the logs. The first culprit was error message on SPIFFS mount fail. So... partition... Surely, it was first time to hear SPIFFS so did a bit of reading on ESP32 partitions and file systems. After struggling a bit, found that the cause is SPIFFS partition is not formatted. By checking the E-SPIFFS init code it doesn't have the 'format if fails' flag set. Hacked the library init code to force the initial format. Once it's formatted the error message was gone.

Challenge 2
The mount error is gone but Ignitron was still resetting on BT connection. Paying another attention finds fail on reading default preset. Read the last part of software section where ESP32 Sketch Upload Manager is mentioned. Got Arduino IDE 1.8.x, force to run it with Windows 7 compatible mode(Windows 11 Pro here). Got the add-in installed, loaded the Ignitron.ino and downloaded the data folder. It reports all good. Booting Igintron still shows the same error. Started to debug from code level to project level settings. Tested pure SPIFFS code and it just worked fine. Write and read file work both on Arduino IDE and PlatformIO but whatever I do the Ignitron fails to read the file so thought it must be something related with project level. Finally found that it was the partition layout difference between PlatformIO and Arduino IDE as I was not using node32s board. Checked the project and saw Ignitron is using no_ota.csv which matches with No OTA (2MB APP/2MB SPIFF) setting for ESP32 Dev Module. Only if there was a warning on the instruction. But no blame here. It way my short of knowledge on ESP32 as I said, it was my first try. Actually it was a good chance to understand ESP32 partitioning better.

The boy spent some time to design a 3D model of the case, printed it out, then all assembled. I just did all direct spaghetti wire soldering for the sake of time saving for the PCB and part order. We can share the 3D print model if anyone wants.

By having all assembled the boy spent a bit of time to play and seemed pretty happy until.....

He found frequent switching between effects crashes the Spark Mini amp(not Ignitron). Had to hard power reset on the amp when it happens. I may need to flash the debug enabled firmware, hook up to the serial monitor and see what it shows but my gut feeling is that the frequent change screws up the BT communication and it makes the amp crashes.

Anyway here is what we have today and again thanks guys for all you have done on this project. You are AWESOME!

[stangreg]

@intaker Thanks for sharing your experience with building Ignitron. This is very valuable as it provied feedback to me how I can improve the documentation. I also like the case design your son has made!
Let me share some comments to your story:
IDE - I have just recently switched to VS Code and PlatformIO. It is very likely that some of the documentation still refers to my old way of developing (using Sloeber and Arduino IDE 1.8). After switching to VS Code, I am now able to do everything in the same IDE.
SPIFFS - based on the past feedback I received I know that this is a pitfall for some users. As it is a one-off task to get the partition right and upload the data, I did not spend too much time on figuring out what to consider. However, I have never required to mess with the SPIFFS code. I have used the targets coming with the Platform IO plugin (Build Filesystem image, Upload Filesystem Image) which worked after a bit of configuring.
PlatformIO.ini - As I have only an ESP32, I have updated the platform.ini file with relevant settings for that platform. If you are using another platform, I assume you can just reuse some of the ESP32 settings and put them in the other platform config part.
Spark crashes - I have also noticed this, I assume this happens when too much data is sent to the Spark at once (quick changes). Maybe the Spark amp is running out of memory, don't know. Also sometimes the BLE connection is not stable, here I don't know if it is Spark Amp, the NimBLE library or my configuration. After a restart usually it works fine. Maybe you want to regularly check for code updates, I have recently changed a bit around the communication which hopefully makes it more stable. I am currently working on integrating the looper functionality of Spark 2 amp into Ignitron.

Regarding your comment of spending a few evenings, I have actually spent a lot of time over the past years into the code to extend it with more functionality and to support more models from Spark. It is really a fun project. I started this before the official pedals for Spark were published and am a bit proud that Ignitron adds more functionality, e.g. around custom presets, controlling external apps via BLE keyboard etc.

I hope everything works fine for you now. Does the AMP mode work now for you?

[intaker]

@stangreg Thanks for your comments and yes the AMP mode works fine. I haven't tried Platform IO plugin (Build Filesystem image, Upload Filesystem Image) but sounds like a right direction as everything can be done in VS Code. I will give a go to extend my knowledge. The looper integration sounds very interesting. My son was trying to run a looper on his phone but not many phone looper apps support external bluetooth keyboard. Will follow the project and report back from my tries as well. I flashed a debug enabled firmware on a prototype board last night and asked my son to try. We could see the crash so planned to get some logs next week. And... no wonder for your time spent on the project. I could see your efforts through out the codes and appreciate it.

[stangreg]

@intaker Regarding the looper app, I am using Loopy HD, but that seems to be available only for iOS. A quick search by me for an Android app which can be controlled via Keyboard was not successful.

[intaker]

Thanks @stangreg for sharing the info

[intaker]

Getting some logs.

• Test was toggling effect using Ignitron. The AMP itself changes the effect but Igintron doesn't display the change. Log shows empty Unstructured data: followed by 0315 not handled. In below log, it stops working at message 70 and didn't for the the following messages
• Spamming effects causes the problem

01FE000053FE2E000000000000000000F001420B01150211314B6C6F6E4300656E746175725320696C76657242F7
COMPLETE MESSAGE: 
01FE000053FE2E000000000000000000F001420B01150211314B6C6F6E4300656E746175725320696C76657242F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE2E000000000000000000F001420B01150211314B6C6F6E4300656E746175725320696C76657242F7
Done with this command!
Button pressed: 25
Switching Off effect KlonCentaurSilver...MESSAGE NUMBER: 67

01FE000053FE2E000000000000000000F001430B01150211314B6C6F6E4300656E746175725320696C76657242F7
COMPLETE MESSAGE:
01FE000053FE2E000000000000000000F001430B01150211314B6C6F6E4300656E746175725320696C76657242F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE2E000000000000000000F001430B01150211314B6C6F6E4300656E746175725320696C76657242F7
Done with this command!
Button pressed: 26
Switching On effect Phaser...MESSAGE NUMBER: 68

01FE000053FE22000000000000000000F001445E01150206265068617365027243F7
COMPLETE MESSAGE:
01FE000053FE22000000000000000000F001445E01150206265068617365027243F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE22000000000000000000F001445E01150206265068617365027243F7
Done with this command!
Button pressed: 32
Switching On effect DelayMono...MESSAGE NUMBER: 69

01FE000053FE25000000000000000000F0014507011502092944656C6179104D6F6E6F43F7
COMPLETE MESSAGE:
01FE000053FE25000000000000000000F0014507011502092944656C6179104D6F6E6F43F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE25000000000000000000F0014507011502092944656C6179104D6F6E6F43F7
Done with this command!
Incoming block: F0015407031502092944656C6179104D6F6E6F43
Incoming block: 00F7
Processing block
F0015407031502092944656C6179104D6F6E6F43
Processing block
00F7
Response so far:
F0015407031502092944656C6179104D6F6E6F4300F7
Unstructured data:

0315 not handled
09A944656C61794D6F6E6FC300
Last message number: 54
Button pressed: 33
Switching Off effect bias.reverb...MESSAGE NUMBER: 70

01FE000053FE27000000000000000000F00146030115020B2B626961732E4072657665726242F7
COMPLETE MESSAGE:
01FE000053FE27000000000000000000F00146030115020B2B626961732E4072657665726242F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE27000000000000000000F00146030115020B2B626961732E4072657665726242F7
Done with this command!
Incoming block: F00155030315020B2B626961732E407265766572
Incoming block: 62420000F7
Processing block
F00155030315020B2B626961732E407265766572
Processing block
62420000F7
Response so far:
F00155030315020B2B626961732E40726576657262420000F7
Unstructured data:

0315 not handled
0BAB626961732E726576657262C200
Last message number: 55
Button pressed: 33
Switching Off effect bias.reverb...MESSAGE NUMBER: 71

01FE000053FE27000000000000000000F00147030115020B2B626961732E4072657665726242F7
COMPLETE MESSAGE:
01FE000053FE27000000000000000000F00147030115020B2B626961732E4072657665726242F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE27000000000000000000F00147030115020B2B626961732E4072657665726242F7
Done with this command!
Button pressed: 26
Switching On effect Phaser...MESSAGE NUMBER: 72

01FE000053FE22000000000000000000F001485E01150206265068617365027243F7
COMPLETE MESSAGE:
01FE000053FE22000000000000000000F001485E01150206265068617365027243F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE22000000000000000000F001485E01150206265068617365027243F7
Done with this command!
Button pressed: 25
Switching Off effect KlonCentaurSilver...MESSAGE NUMBER: 73

01FE000053FE2E000000000000000000F001490B01150211314B6C6F6E4300656E746175725320696C76657242F7
COMPLETE MESSAGE:
01FE000053FE2E000000000000000000F001490B01150211314B6C6F6E4300656E746175725320696C76657242F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE2E000000000000000000F001490B01150211314B6C6F6E4300656E746175725320696C76657242F7
Done with this command!
Button pressed: 32
Switching On effect DelayMono...MESSAGE NUMBER: 74

• There is a specific custom preset(Say it) which shows the issue immediately for button 1 and 2. May be different reason because the log doesn't show 0315 not handled but some garbage data. When that happens pressing any other button fixes the issue.

JSON:
{"PresetNumber": 127, "UUID": "c56c513d-3288-4060-83da-3f8d5ca0eb74",
"Name": "Say it (itte)", "Version": "0.7", "Description": "", "Icon": "icon.png", "BPM": 120.0000, 
"Pedals": [
{"Name": "bias.noisegate", "IsOn": true, "Parameters":[0.0485, 0.7474, 1.0000]},
{"Name": "Compressor", "IsOn": false, "Parameters":[0.3029, 0.5233]},
{"Name": "Overdrive", "IsOn": true, "Parameters":[0.6616, 0.7000, 0.5300]}, 
{"Name": "Checkmate", "IsOn": true, "Parameters":[0.9385, 0.6586, 0.4289, 0.3258, 0.4387]},
{"Name": "GuitarEQ6", "IsOn": true, "Parameters":[0.5000, 0.0000, 0.3683, 0.5000, 0.5000, 0.5000, 0.5000]}, 
{"Name": "DelayRe201", "IsOn": false, "Parameters":[0.2337, 0.1005, 0.4729, 0.3109, 1.0000]},
{"Name": "bias.reverb", "IsOn": true, "Parameters":[0.7370, 0.1500, 0.5000, 0.6214, 0.0357, 0.3500, 0.1000, 1.0000]}],
"Filler": "46"
}

ACK number 10
Acknowledgment for command 04 38
Last message number: 0A
Received final ACK
Received ACK for 0x38 command
OK
Button pressed: 25
Switching Off effect Overdrive...MESSAGE NUMBER: 11

01FE000053FE25000000000000000000F0010B3201150209294F76657264107269766542F7
COMPLETE MESSAGE:
01FE000053FE25000000000000000000F0010B3201150209294F76657264107269766542F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE25000000000000000000F0010B3201150209294F76657264107269766542F7
Done with this command!
����������������������������������������������������������������pressed: 18
Switching to FX mode
Re-initializing button config
Button pressed: 25
Switching Off effect Overdrive...MESSAGE NUMBER: 12

01FE000053FE25000000000000000000F0010C3201150209294F76657264107269766542F7
COMPLETE MESSAGE:
01FE000053FE25000000000000000000F0010C3201150209294F76657264107269766542F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE25000000000000000000F0010C3201150209294F76657264107269766542F7
Done with this command!
����������������������������������������������������������������ressed: 25
Switching Off effect Overdrive...MESSAGE NUMBER: 13

01FE000053FE25000000000000000000F0010D3201150209294F76657264107269766542F7
COMPLETE MESSAGE:
01FE000053FE25000000000000000000F0010D3201150209294F76657264107269766542F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE25000000000000000000F0010D3201150209294F76657264107269766542F7
Done with this command!
Button pressed: 26
Switching Off effect GuitarEQ6...MESSAGE NUMBER: 14

01FE000053FE25000000000000000000F0010E6E01150209294775697461107245513642F7
COMPLETE MESSAGE: 
01FE000053FE25000000000000000000F0010E6E01150209294775697461107245513642F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE25000000000000000000F0010E6E01150209294775697461107245513642F7
Done with this command!
Button pressed: 32
Switching On effect DelayRe201...MESSAGE NUMBER: 15

01FE000053FE26000000000000000000F0010F100115020A2A44656C617920526532303143F7
COMPLETE MESSAGE:
01FE000053FE26000000000000000000F0010F100115020A2A44656C617920526532303143F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE26000000000000000000F0010F100115020A2A44656C617920526532303143F7
Done with this command!
Incoming block: F0010F000415F7
Processing block
F0010F000415F7
Response so far:
F0010F000415F7
Unstructured data:

ACK number 15
Acknowledgment for command 04 15
Last message number: 0F
Received final ACK
OK
Button pressed: 32
Switching Off effect DelayRe201...MESSAGE NUMBER: 16

01FE000053FE26000000000000000000F00110110115020A2A44656C617920526532303142F7
COMPLETE MESSAGE:
01FE000053FE26000000000000000000F00110110115020A2A44656C617920526532303142F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE26000000000000000000F00110110115020A2A44656C617920526532303142F7
Done with this command!
Incoming block: F00110000415F7
Processing block
F00110000415F7
Response so far:
F00110000415F7
Unstructured data:

ACK number 16
Acknowledgment for command 04 15
Last message number: 10
Received final ACK
OK
Button pressed: 32
Switching On effect DelayRe201...MESSAGE NUMBER: 17

01FE000053FE26000000000000000000F00111100115020A2A44656C617920526532303143F7
COMPLETE MESSAGE:
01FE000053FE26000000000000000000F00111100115020A2A44656C617920526532303143F7
Sending message via BT.
Sending message via BT.
Connection ok
Sending block:01FE000053FE26000000000000000000F00111100115020A2A44656C617920526532303143F7
Done with this command!
Incoming block: F00111000415F7
Processing block
F00111000415F7
Response so far: 
F00111000415F7
Unstructured data:

ACK number 17
Acknowledgment for command 04 15
Last message number: 11
Received final ACK
OK

Was not able to crash the amp yet. Will try later more

[intaker]

Flashed branch 20240924_read_effect_onOff and a quick test shows it has fixed the issue(at least on Ignitron display side) which is a very good sign. The boy will test the build for a couple of days and report the result. Thanks a lot for the quick test code!

Regarding your comment on same effect on different presets, I wonder whether it can be overload on the amp side. We only have Mini so can't really test with a more capable amp.

[intaker]

@stangreg The boy has been using the branch build for a few days and very happy with it. I can confirmed that the effect switching works as expected and believe the branch can be merged. Thanks again for the fix!

[stangreg]

@intaker Tranks for the positive response! I am glad that everything works fine. I have now merged the branch into main in combination with some bigger changes to enable controls for Spark 2's built-in looper.

[intaker]

@stangreg, a few updates from our side

1. Got the latest main and confirmed it works okay
2. Tested the preset file system download using PlatformIO. Much better than old Arduino IDE
3. Found a case where Spark Mini crashes
4. Start to see a crash I didn't see before in the serial log

Attachment: Ignitron_local.patch

Details here

1. Got the latest main, merged my changes for the display then it worked okay. I wonder whether there will be a nice way to support both SD1306 and SH110x through the source level #define. (Code change in the attached patch file)
2. Using PlatformIO file system is easy and effective as it will respect the partition file. I also saw that you have updated the documentation. Thanks for the update
3. My son was testing a preset by preset and found a specific preset crashes his amp.
   • After a few trials and errors we found that it's the length of the preset name. The symptom is changing to the preset is okay. BT message sent to the amp. Seems amp changed to the preset but any action from either Ignitron or the amp crashes the amp(frozen)
   • Tested with a few lengths and we found Mini only supports up to 31 characters for the name. I guess it's 32bytes with last charter to be the null character(\0) to terminate the string. It could be a limitation of Mini only but I can't test as we don't have another amp.
   • Checked all presets and found some with longer name. Also found some untidy preset names and fixed them(Code change in the attached patch file)
   • I was going to change the preset json files with shorter names when my son suggested a different approach. Just trim the name in the BT message to amp so Ignitron can display the full name and the amp doesn't crash. Took the approach and it worked!(Code change in the attached patch file)
4. I start to see a heap error. Seems related with memory free. I see this without my changes so I'm confident it's not related with my changes at least

"Filler": "37"
}
Changing to preset 01-1...OK
Changing to preset 01-2...Changing to preset 01-4...OK
OK
Changing to preset 14-4...OK
Changing to preset 14-3...OK
Changing to preset 14-4...OK
Characteristic FFC1 not found.

assert failed: heap_caps_free heap_caps.c:381 (heap != NULL && "free() target pointer is outside heap areas")


Backtrace: 0x40083e9d:0x3ffd8ef0 0x4009508d:0x3ffd8f10 0x4009ae71:0x3ffd8f30 0x400842e2:0x3ffd9060 0x4009aea1:0x3ffd9080 0x401ba441:0x3ffd90a0 0x400d2daa:0x3ffd90c0 0x400d74b6:0x3ffd90e0 0x400d7fb3:0x3ffd9130 0x400d878e:0x3ffd9150 0x400d87d0:0x3ffd9180
  #0  0x40083e9d in panic_abort at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/esp_system/panic.c:408
  #1  0x4009508d in esp_system_abort at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/esp_system/esp_system.c:137
  #2  0x4009ae71 in __assert_func at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/newlib/assert.c:85
  #3  0x400842e2 in heap_caps_free at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/heap/heap_caps.c:381
      (inlined by) heap_caps_free at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/heap/heap_caps.c:366
  #4  0x4009aea1 in free at /home/runner/work/esp32-arduino-lib-builder/esp32-arduino-lib-builder/esp-idf/components/newlib/heap.c:39
  #5  0x401ba441 in operator delete(void*) at /builds/idf/crosstool-NG/.build/HOST-x86_64-w64-mingw32/xtensa-esp32-elf/src/gcc/libstdc++-v3/libsupc++/del_op.cc:49
  #6  0x400d2daa in __gnu_cxx::new_allocator<unsigned char>::deallocate(unsigned char*, unsigned int) at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\ext/new_allocator.h:125
      (inlined by) std::allocator_traits<std::allocator<unsigned char> >::deallocate(std::allocator<unsigned char>&, unsigned char*, unsigned int) at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/alloc_traits.h:462
      (inlined by) std::_Vector_base<unsigned char, std::allocator<unsigned char> >::_M_deallocate(unsigned char*, unsigned int) at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/stl_vector.h:304
      (inlined by) std::_Vector_base<unsigned char, std::allocator<unsigned char> >::~_Vector_base() at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/stl_vector.h:285
  #7  0x400d74b6 in std::vector<unsigned char, std::allocator<unsigned char> >::~vector() at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/stl_vector.h:570
      (inlined by) CmdData::~CmdData() at src/SparkTypes.h:102
      (inlined by) void __gnu_cxx::new_allocator<CmdData>::destroy<CmdData>(CmdData*) at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\ext/new_allocator.h:140
      (inlined by) void std::allocator_traits<std::allocator<CmdData> >::destroy<CmdData>(std::allocator<CmdData>&, CmdData*) at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/alloc_traits.h:487
      (inlined by) std::deque<CmdData, std::allocator<CmdData> >::pop_front() at c:\users\keith.lee\.platformio\packages\toolchain-xtensa-esp32\xtensa-esp32-elf\include\c++\8.4.0\bits/stl_deque.h:1594
      (inlined by) SparkDataControl::sendNextRequest() at src/SparkDataControl.cpp:913
  #8  0x400d7fb3 in SparkDataControl::triggerCommand(std::vector<CmdData, std::allocator<CmdData> >&) at src/SparkDataControl.cpp:897
  #9  0x400d878e in SparkDataControl::readHWPreset(int) at src/SparkDataControl.cpp:1222
  #10 0x400d87d0 in SparkDataControl::checkForMissingPresets(void*) at src/SparkDataControl.cpp:1250





ELF file SHA256: 0beda133f3d4e243

Rebooting...
ets Jul 29 2019 12:21:46

rst:0xc (SW_CPU_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:2
load:0x3fff0030,len:1184
load:0x40078000,len:13232
load:0x40080400,len:3028
entry 0x400805e4

Now my son is using the latest version and everything seems working fine.
Thank you!

[stangreg]

@intaker Thanks for the valuable information! I will look into this in a few days when I find some time. I have some idea already on the heap one.
On the long name, it might work when I do the same as for the description, I just need to encode the string differently when it is longer than 31 characters. Well spotted!

[barsk]

Great work everyone! Much appreciated.

I will look into building an Ignitron myself, having some Arduino and PIC experience. ESP32 will be a new for me.

@intaker: I would also like to 3D print the case which looked very nice in your version here. Would you like to share it perhaps? If possible perhaps the editable CAD version if I see something I would like to improve on? I use Fusion 360 and it can import a heap of different file types.

[barsk]

Ok, time to assemble. The 3D design has went through some revisions already, mostly with tolerances that was a bit too tight at places.
However as I got the PCB just the other day and was looking at how to asseble it dawned on me that the PCB is not compatible at all with the 3D design. The PCB has the resistors overhanging the USB port, making it impossible to mount the assembly flush with the 3D case wall. That makes a 6mm gap unfortunately.
I have to either live with the huge gap, cut of the resistor part of the PCB with a dremel and solder in the resistors in an "ugly" way - or make my own PCB, possibly with a standard prototyping breadboard type PCB.

[intaker]

@barsk, I feel I should have told you that we didn't use any PCB but direct soldering intentionally due to the design my son wants. He wants a replaceable outer case so he can print different outer cases(different colours, extended to hold batteries, and etc...) and quickly drop in the main body. He also wanted to use the stencil letters with LED behind. The ideas didn't allow to have enough space to host the PCB while keeping the overall size slim. The approach also required precise alignment of esp32, OLED and LEDs to flush to the surface, and some hot gluing. As a result the case became very customised one. I found myself prefer direct soldering as I just need to make one and it also gives a sort of retro look like an old hand-made amp😄 but as I guess you don't need the replaceable skin, I would suggest to get more space by making the external walls thinner and rearranging internal walls then use the PCB. Here are some photos of the final build just in case.

[barsk]

Hi @intaker. No big deal!
I'll work around the issue, I think just cutting away the inner "main" parts, and minimizing the wall thickness of the casing will help a lot. The USB port will then be reachable as mounted on the PCB. If that becomes impractical, I'll roll my own PCB either from a prototype board or from a ECAD design proper. I'm a bit rusty in that department, but I had a working solution with using laserjer toner as masking for a PCB and then just etching away. Gets a really nice PCB, but it has been some years since I did that last. Memory fades... :)

[intaker]

@barsk I've been dreaming of making custom PCBs but never followed the dream. laserjer toner as masking for a PCB and then just etching away sounds very tempting! Just googled and saw a few sites. I may finally give a go 😄

[barsk]

It is really useful! tI involves a bit of trial and error though until you have perfected the process which is is dependent on the type of laserjet you have. I also have a "t-shirt" press for transferring vinyl for clothes cut out on a vinyl cutter. I use that as my heat source that transfers the toner to the copper. You have to print the PCB wires and holes from a gerber file converted to a printable file in 1:0 scale and mirrorred. I had some problems figuring out which layers from the PCB design to use and the actual tool to create the printable file in high res. That is the part that I may have not documented and where memory have faded away. If you don't have t-shirt press (who has? :) ) then a hot iron will do nicely. The toner must be printed on a glossy paper that can release the toner when heat is applied. Use printer settings to maximise toner thickness (contrast, blackness, paper type). I ended up using the backing paper of stickers or vinyl (that I use for my vinyl cutter anyway). Those had the right properties. There are other such papers on ebay and Amazon, and i bought some but I had no luck with at least with the brand I found.
Once you have perfected the toner transfer there is no need for exposure/developing etc. It is just the final step in the PCB process - etching. If the toner is covering enough, you will have some great results. I could even make some fairly nice SMD PCB's and even double sided. You just have to line up the holes from both sides. I did that by drilling two holes from in each corner of the pcb corresponding to actual component holes (with a paper mask). Then matching up the pattern on both sides. Same here though, the exact procedure is a bit forgotten. To connect between layers (vias), you can just have a wire through and solder on both sides. There actually is "vias" you can buy in a little box and push through and fasten with a hit of a small hammer and a socket tool. That is a bit extra though. :)

[intaker]

@stangreg, Seems my son's Ignitron has a dedicated spot in his gig bag and he told me it's very useful. He comes with one possible improvement idea.
Problem : When Ignitron connects to an amp, it defaults to H/W Preset #1 regardless what is the current preset of the amp. This is not a problem in normal use cases but it causes a sudden preset change when Ignitron resets unexpectedly during a live session
Possible solution 1(preferred) : Ignitron reads the amp's current preset on connection and matches to it
Possible solution 2 : Ignitron connects but waits on user input rather than changing the amp's preset automatically
Thanks.

[intaker]

Thanks @stangreg for considering it. Just let us know if there is anything we can assist.

[stangreg]

@intaker @barsk I have now finished the latest changes to the code. Ignitron will now load the current preset from the amp after connecting and show it in the display. I have also done some refactoring in the structure of the code and did some stability/performance adjustments. Feel free to load the latest code from the main branch and try it out. Feedback is welcome!

[barsk]

Nice. I will give it a try soon. I caught the flu and been a bit under the weather for a while. I'm going to order the components and PCB which probably takes a while to arrive as well.

[barsk]

I have come a bit further now. I've got the hardware and the ESP32 dev module (38 pin) and got VS Code + Platform.IO up and running. I'm just missing the PCB which should come in a few days.
I started with the firmware though and ran into the same problem as @intaker. SPIFFS mount failure. I added the line:
SPIFFS.begin(true); in the setup() method just below the first Serial.print(). It seems to have solved the issue. Why not have that line there always? The only (?) situation data could be erased is if the partition table is changed which is unlikely?
FAQ seems to support this: https://docs.espressif.com/projects/arduino-esp32/en/latest/troubleshooting.html?highlight=spiffs#spiffs-mount-failed

[stangreg]

@barsk Good point, I have added the SPIFFS.begin(true); to the intialization of the PresetBuilder before it tries to setup the presets from the filesystem.
BTW: @intaker @barsk The latest version in main now includes a tuner function to display the tuner output of the amp, so you can quickly tune between songs without using an external tuner.

[intaker]

That's a great news @stangreg, I will give a go and share. Thanks.

[intaker]

Hi @stangreg, the boy finally came back from a trip so we gave a go with the latest version. It's just fantastic. The preset reading works like a charm. We both were a bit confused when HW 0 appeared but quickly realised that it's an initial landing zone. A smart approach! Tested with both amp and Ignitron reset scenarios and confirmed that it does work.
The tuner is very impressive. Really handy as a tuner and muter. The OLED display design is a top job. I love the full screen invert and dynamic arrows. Thanks heaps for all the improvements.

[barsk]

I'd like to tune in with the cheering here. As I now have finally got everything together and being able to try the functionality out I am just left with a deep awesomeness feeling. Wow, wow, wow what a tool you have created here! It works just splendidly, and so easy to switch between tons of banks of different tones, modifying each in FX mode if required. Just awesome. Thanks!!!

I have done some (a lot) of modifications on @intaker 3D CAD design and although it looks similar to the old design, a lot of minor changes have been made. Need to make some final adjustments though as the height was just slightly to low to allow for the PCB connectors and wires. When reprinted and reassembled I will take some pictures. I have also done some code changes to easily handle the SH1106 driver for the 1.3" OLED. Bit rusty on git, but I'll try to make some patches for you when time allows. The driver API is almost identical, but the object handles are different as well as the WHITE and BLACK defines.