We measured the deviation limits on the MD-380 with firmware from OpenRTX. Thank you to Redman for help with modifying the firmware to make this test as easy as possible.
The transmitted signal is about 10 dB down at 3000 Hz and almost gone at 4200 Hz. Therefore, there is not enough deviation for Opulent Voice.
The part of the radio under test was the HR_C5000. This is a part from Hong Rui and is a DMR digital communications chip. The chip handles 4FSK modulation and demodulation, among other functions.
According to a translated datasheet for the HR C5000, adjustment of the frequency offset range not possible. It appears to be designed only for +/- 3 kHz. Unless there’s an undocumented feature, or gain is added after the HR C5000, then +/- 3 kHz is the maximum deviation for this radio.
The HR C5000 puts out two analog signals MOD1 and MOD2, which are combined and then drive a varactor diode. The varactor might well have more range. Or it could be replaced with one that has more range.
Below is a photo essay of the testing and screenshots of results.
30 AWG wire soldered to pin 5 of the HR_C5000 on the MD-380. This is one of two audio inputs to the modulator. It’s the one used for signaling tones, and not the one used for microphone audio, in the original design. Ground wire attached as well and brought out. Firmware modifications disabled M17 baseband output when PTT pressed. Pre-emphasis and filters disabled by putting radio in M17 mode. The red wire is part of the standard M17 mod for the MD-380.Ground wire attached to point on board and brought out. Modifications inspected by KB5MU in Remote Lab. The black rubbery weather seal gasket around the perimeter of the cast heatsink means there’s no way for wires, even very skinny ones, to come out through the seam in the case.Hole was drilled in the side of the case and the wires brought out. Radio was put back together. Battery, display, etc. all working after modifications. Notice the 20 dB attenuator on the output of the HT. Its lowest output power is 1W nominal, which matches the maximum rated input of the spectrum analyzer, so the attenuator was added to protect the spectrum analyzer’s front end. Test setup.
Test Results
Transmitted signal before modified firmware.Transmitted signal after modified firmware. Notice no modulation because the baseband signal has been disabled. 1kHz sine wave at 100 mV injected to pin 5 of the HR C5000.1kHz sine wave at 220 mV injected to pin 5 of the HR C5000.3 kHz sine wave at 220 mV injected to pin 5 of the HR C5000. Signal is about 10dB down at 3 kHz4200 Hz sine wave at 220 mV injected to pin 5 of the HR C5000. Signal is approximately 45 dB down. Looking at the low end, this is one half a Hz at 220 mV injected to pin 5 of the HR C5000.Two tenths of a Hz at 220 mV injected to pin 5 of the HR C5000. Notice the discontinuities. These were not seen when the signal generator was connected to an oscilloscope. It could be that when the modulating frequency is too low, it interacts with a PLL or other frequency stabilization loop in the radio. It would not be an issue in the original design, with all audio coupled through capacitors. No way anything that close to DC would get to the audio chip.
Excellent. Thank you very much for sharing.