About whether the software needs to set dio2 of sx1262 in different circuits

[weekroom]

At present, the IRQ of my ESP32 is connected to DIO1. Under the following two circuit designs, should the software be set during initialization setDio2AsRfSwitch(true)

[jgromes]

I don't know how those two schematics are related to each other. The point of DIO2 is to make RF Rx/Tx switching easier, it will automatically trigger its state based on whether the SX1262 is transmitting, or receiving. This is also the default configuration in RadioLib.

The RF switch in your schematic has complementary control inputs, CTRL and /CTRL, so I guess the idea of the transistor is to invert the DIO2 signal level. That can work (unless you want to, e.g. power down the RF switch). But I don't know if the inverter output RXEN is connected to the ANT_SW input.

[weekroom]

At present, I found no dio2 circuit in the E22-900MM22S module that I got. Figure 2 is designed by the hardware engineer according to the circuit in Figure 1. I used the instrument to check and found that it can trigger its state automatically according to whether SX1262 sends or receives, but the actual transmission distance is not far and unstable. If the road design in Figure 1 is used, the transmission distance can be very long and stable, is there any special difference between the two

[jgromes]

FWIW I have previously used this simple transistor switch to control both RFEN and TXEN with just DIO2.

[weekroom]

I checked the Settings, through the hardware TXEN, RXEN high and low level reading, DIO2 can switch to receive and send normally, but it is unstable, or the transmission distance is short

[lyusupov]

I use one E22-900M22S since 2019
It has RF switch IC built-in

Circuit designer who is going to operate with E22-900MM22S has an option to either

• let the SX1262 to drive RF switch by DIO2 (most common use case), or
• drive RXEN and TXEN by an MCU

Important

I can see that your intention is to use E22-900MM22S (two M in the middle)
Internal hardware design of the E22-900MM22S may differ from E22-900M22S

Ebyte most likely uses PE4259 in many of own module designs.

Logic of this IC is explained in this thread: #1320

[lyusupov]

Thank you so much for your help. Thank you so much

[weekroom]

At present, I am testing at a distance of one kilometer, and sometimes I can receive continuous, intermittent or not receive at all. I think it is also a problem of unstable communication of 404 2220, which can be sent and received normally when using 100 but I do not have enough evidence to prove it. I have asked the technical support of ebyte. They told me that 404 2220 was only changed to 100 package because of the price increase, they do not admit that there is a problem with this one, is there any way to troubleshoot this problem

[weekroom]

One method has been tested. One device hardware puts TXEN high and RXEN low, enabling it to send data in send mode; The other device hardware has TXEN low and RXEN high to receive data in receive mode. This method allows the equipment to communicate stably for up to a kilometer

[HeadBoffin]

If you are building a product, I'd test a range of vendors modules. If you are building a one off device, maybe simpler to switch to a different vendor rather than spend a lot of time resolving something that could have batch issues in the future.

[weekroom]

At present, after I replace 404 2220 with CKRF2179MM26-C4 device of silk screen 100, it will become much more stable than 404 2220