After a dozen flights without a glitch on my new sparkie over the last month, I had two flyaways in one day with sparkie. The first dozen flights were back east and when I returned to the west coast, the first thing I did was recalibrate the IMU and compass. All looked fine. Took sparkie out to my local flying field to get a couple of flights. I set a distance limit of 200 meters...which is as far as I can see it, and a hight limit of 50 meters. Took off...did stability checks...all fine...started my planned route to photograph our runway, pilot area and some of the surrounding area. Within a few minutes, I noticed the speed has increased and the distance getting farther heading east. Good thing it's about 5 km of flat grassland. I though ok...it'll stop at the 200 meter distance...nope...I noticed that it was in ATTI mode and was approaching 300 meters. I saw it switch to GPS and hit the RTH and it worked...sparkie returned to visual range and I was able to take over and manually land it. Heart rate went down to normal. Ok...I said to myself...must have pooched to calibration. So calibrated IMU and compass again. Looked good....no apparent problem, took off again, deciding to keep it within the confines of the airfield (MAAC airfield btw) and all seemed good...going over to the far side and at 30 meters, once agiain it flipped to ATTI mode and started heading north. Within a seconds I saw it flip to GPS mode and I did the RTH. That was enough of sparkie flights for that day until I figure what is going on. Retrieved the DJIGO 4 logs for both flights and uploaded them for analysis. In both flights...are roughly 30 meters and over the same patch of ground, the log shows electromagnetic interference, compass error and yaw error. It seems rather coincidental that there was strong electromagnetic interference in the logs at about the same spot and altitude. I am wondering if the interference could be caused by high power microwave and cell towers. There have been some new ones installed within a mile or so, but I would have thought the spark would be shielded from that...maybe not. Any opinions? What about the software, anyone have an comments on whether the DJIGO 4 or the DJI pilot app should be used for calibration? Until these flights, I had zero problems with the DJI go4 app. Thanks.