Our LM2842 is used to supply a power amplifier to communicate in FSK. The output of the power amplifier is protected by external clamping diodes.When a parasitic signal is received, it is demodulated by the clamping diodes and result into a parasitic DC voltage on the power supply rail powered by the LM2842.
This prevents the LM2842 to start when the power amplifier is switched on.
So, the LM2842 does not start when the following conditions are met:
- LM2842 in stand-y as a consequence of a very low current on its output: 0.1mA
- AND a DC voltage on its output coming from an external source with about 1k ohms impedance
- AND a sudden current step (when our amplifier is switched on) up to 65mA
If we add a load on the output of the LM2842 (2k ohms) we partially solve the problem as the LM2842 is always active in this case. However if the noise is too strong, this is not enough...
We compared the design of the LM2842 with a TI LMR12007, the difference from an outside point of view is an additional diode on the BOOST pin. An additional zener diode is also added in some conditions.
Is the starting problem of the LM2842 linked to this missing diode ?