Rethinking encoder controls

[max-arnold]

The current encoder controls can be summarized as follows:

• Single (short) click - triggered by the trailing edge
• Middle press (half a second) - triggered by the trailing edge
• Long press (2 seconds) - triggered by the time interval (no need for a user to guess "am I pressing the button long enough and is it ok to release it?")
• Rotate
• Press + rotate

The click and rotate do the obvious things. The middle press is used as a volume control shortcut. The long press is used to disable/enable the screen. Press + rotate is used to trigger the auto-tuning process on AM/FM, or tuning with a larger predefined step on SSB.

The latter seems to be not so convenient. A proposed improvement is the following:

How about the following approach (SSB only): push + rotate will switch to a predefined faster step (as it already does now with the lookup table), but you can stop pushing it and still rotate the knob to tune the frequency using the faster step. After a certain period of no activity the step will switch back to the initial value.

An implicit control logic includes modal modes (like the various menu options) and timeouts.

Also there was a debate in a radio community Telegram chat on how the display sleep should behave:

• Triggered by an explicit action and/or timeout (currently both using the long press and a timeout, the timeout is optional). This seems to work well, but the long press can't be used for any other actions
• The encoder rotation and short clicks should be locked, to prevent detuning while the receiver stays in a pocket (the current behavior, but rotation or short click will wake it up if the timeout is enabled, otherwise it requires a long press)
• Rotation and clicks should be allowed to enable blind tuning and volume control while listening for weak signals (a disabled display measurably reduces QRM)
• Rotation should control the volume

One proposed solution that allows both the locked state and blind tuning is to lock the encoder after the screen was locked for more than 3 seconds. If there was any movement within these 3 seconds, then the encoder won't be locked at all. The second solution is a Settings option that will define the desired behavior.

I also plan to experiment with light or deep sleep modes in ESP32 to save even more energy. It could be combined with the display sleep mode, but doesn't allow the blind tuning scenario.

Another idea is to use middle press or some other shortcut like press + rotate to define the current step. This could be a shortcut to the step menu, or a frequency digit selection like on my FX-4CR transceiver:

This nicely handles the steps that are multiples of 5, and allows to drop the step menu completely, but doesn't handle the 9kHz step.

Alternatively, this can be used as a direct frequency input method.

Cc: @G8PTN

[G8PTN]

Hi Max,

Another idea is to use middle press or some other shortcut like press + rotate to define the current step. This could be a shortcut to the step menu, or a frequency digit selection like on my FX-4CR transceiver:

Yes, using the marker under the digits is an option for step selection and used on many of the touch screen transceivers. Effectively, it allows direct entry.

I also plan to experiment with light or deep sleep modes in ESP32 to save even more energy. It could be combined with the display sleep mode, but doesn't allow the blind tuning scenario.

This would be good to extend the battery life whilst listening.

73, Dave

[G8PTN]

Hi Max,
Just another possible idea. Rather than guessing the time to hold the button, show the “shortcut” option on the display.

• Short press and release
  Normal Menu select

• Press + hold and release
  Shortcut: Volume (Index = 1)
  Shortcut: Band (Index = 2)
  Shortcut: Step (Index = 3)

The Press + hold basically uses a counter incrementing every 1s, creating an index for the shortcut. The shortcut is then displayed on the display and releasing the button actions the shortcut. The example is setup with 3 shortcuts which will just rotate if the button is held, but it could have any number of shortcuts or could be overridden by the manual backlight on/off action. I think with more than a few shortcuts; it is quicker to use the normal menu system.

After the shortcut has been selected, it will automatically become the default for the normal menu select, as per your current firmware.

The timer could probably be reduced, but 1s seems to give plenty of time to release before the next shortcut is presented.

I have shown the shortcut on the status bar, but it could probably be shown differently.

I hacked the ATS_MINI code as a trial and a short video is shown attached. Apologies for the rather poor video.
Code extracts also attached for information.

73, Dave

ATS_MINI.mp4

shortcuts_code_extract.txt

[max-arnold]

Great idea! Unfortunately it makes the long press shortcut that switches the screen on and off impossible.

[max-arnold]

Alternatively, the middle press shortcut could invoke the menu with the default Volume item highlighted (it will require another click to open it though). Then the other menu items will be in predictable positions and can be activated by rotating the encoder.

[G8PTN]

Hi Max,
I think you could still implement the long press backlight. Don’t let the count roll round. In this case the count would stop at a value of 4 and there would be no action.
73, Dave

if (pb1_press_Idx > 0 && pb1_press_Idx < 4){
if (pb1_press_Idx == 1) spr.drawString("Shortcut: Volume",160,12,2);
else if (pb1_press_Idx == 2) spr.drawString("Shortcut: Band",160,12,2);
else if (pb1_press_Idx == 3) spr.drawString("Shortcut: Step",160,12,2);
}

if (pb1_press_inc && (millis() - pb1_press_timer2 >= 1000)) {
pb1_press_timer2 = millis();
if (pb1_press_Idx < 4) pb1_press_Idx ++;
//pb1_press_Idx ++;
//if (pb1_press_Idx > 3) pb1_press_Idx = 1;
drawSprite();
}

[G8PTN]

Hi Max,
I modified the trial code to add long press/hold for display on/off. It uses the function calls that you have for displayOff() and displayOn().

73, Dave

shortcuts_code_extract_2.txt