I made a circuit in which an ISD1760 chip outputs music via an LM386 IC and I like the quality and volume of sound, however, when I run an HM-TRP transmitter on the same circuit board and the transmitter is sending data out at 38kbps, I get the oscillation sound on the speaker (fast motorboat), yet I didn't ask the ISD1760 to play anything. I don't get the sound when the transmitter stops sending data. The transmitter runs on a 915Mhz frequency.

I consulted other forums and I was told its because my gain is too high, so I feel my only options are as follows:

  1. Waste the board, build circuits so the transmitter and audio amplifier don't share the same board. (messy)

  2. Lower the volume of the audio, but if I do that, I might not hear the sound. (not cool)

  3. Make it where the audio and transmissions occur at different times.

Now I may be happy with #3 and I can program a microcontroller to cause #3 to occur but I have two questions on that:

  1. Instead of placing more demands on the micro, could I get away with a bilateral switch and a one-shot timer (like one in the CD4538 cmos IC) and connect the input of the timer to the radio transmit pin so when data is activated, the bilateral switch shuts off the speaker until a byte of data is transmitted and turn the speaker back on?


  1. If idea #3 is doable, will my sound significantly improve or will I still get those oscillations to a worse degree than before?
  • \$\begingroup\$ disconnect the speaker and try again .... it is possible that the speaker picks up the RF directly \$\endgroup\$ – jsotola Jun 5 '18 at 3:39
  • \$\begingroup\$ It's quite possibly a weakness in the PCB layout and/ or decoupling. \$\endgroup\$ – Andy aka Jun 5 '18 at 8:44

RF AM demodulation often occurs due to ESD input diodes or any nonlinearity due to carrier crosstalk . Poor EMC control in the design and lack of crosstalk suppression ground plane or LPF will cause this.

These issues are usually anticipated and corrected by either RF beads on inputs to raise RF impedance and absorb Rf current with input capacitance and/or add suitable value of pF in shunt to nearest ground. Track lengths near 1/4 wavelength on high impedance circuits are even more problematic near RF transmitters and must be corrected by the above due to wave impedance transformations.

Speaker coils are also RF resonant except they tend to force the coil radially causing a rubbing distorted sound so the same treatment as above can be tried. The classic phone tx to tower during idle-connection test sounds like dit-dah-dah-dit repeated for a couple seconds every 5 minutes or so, can be picked up by the speaker or a microphone or any improperly designed LF circuit at >=1GHz.

For example in some cases a 50 ohm "low mu" series SMD bead has the same effect of LPF attenuation as 10 pF shunt, but without specifics , this greatly depends on layout and circuit.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.