5
\$\begingroup\$

I have bought a cheap dashcam to play with, only to discover that my GPS receiver stops working when it's on. After I ruled out power supply issues (both devices have batteries), I'm left with a hypothesis that the dashcam produces EMI interfering with GPS carries frequencies (1.2-1.5 GHz).

Before I start wrapping my head the dashcam in tinfoil, I want to have an actual measurement of the interference I'm facing. Unfortunately, all I have at home is a multimeter and a toy oscilloscope, in no way capable of GHz measurements. I'm thinking about making some sort of antenna probe for my measurements:

schematic

simulate this circuit – Schematic created using CircuitLab

Is this approach realistic? Is there a way to reliably tune it to the frequency range of interest?

\$\endgroup\$
  • 1
    \$\begingroup\$ Reliably tuning things requires a reference you can trust. What you can try is ball-park maths to make an n-winding mini aircoil, preferably from litz or a couple paralleled hair-thin strands of enamelled wire and see if you can get a down-aliased signal from that on a 10x probe. Or rectify it into a tiny cap (1nF or less) and experiment with discharge resistors paralleled, but that requires very quick diodes and a good idea of what you're doing vis-a-vis turns and coil size. To do it with no verification method is just having a few shots in the dark before it gets frustrating. \$\endgroup\$ – Asmyldof Feb 11 '16 at 11:20
  • \$\begingroup\$ Your efforts to measure EMI are pointless. The approach is unrealistic and the "detector" you propose is useless. D1 will not even conduct until you have hundreds of millivolts across C1. You will pick up 1000x less than that at best. Next time: buy better quality products. \$\endgroup\$ – Bimpelrekkie Feb 11 '16 at 11:20
  • \$\begingroup\$ Put as much distance between them as possible: GPS on the left side of the car, Dashcam on the right. \$\endgroup\$ – Steve G Feb 11 '16 at 11:27
  • 3
    \$\begingroup\$ @FakeMoustache: A spectrum analyzer will never show a GPS signal, given that the signal is on the order of 10-20 dB below the thermal noise floor at the Earth's surface. The only way to dig out such a signal is to know a priori what you're looking for (the spreading codes) and integrate the output of a correlator for a significant amount of time (20 ms). \$\endgroup\$ – Dave Tweed Feb 11 '16 at 13:25
  • 1
    \$\begingroup\$ @FakeMoustache Just to be clear, I don't plan to pick up GPS signal. I plan to pick up the EMI signal from the dashcam, which is in the same frequency band. It should be much stronger, especially if I put the antenna a few cm from the source, right? \$\endgroup\$ – Dmitry Grigoryev Feb 11 '16 at 14:00
6
\$\begingroup\$

The best receiver you have is your GPS (or a friend's). Without locking to the spread spectrum code to reduce the effective measurement bandwidth, the signal level is waaay below the ambient thermal noise.

What facilities does your GPS have for showing the strength and quality of the satellite signals? My car Tomtom only tells me how many satellites it has acquired, when it's locked to them. My walking Garmin OTOH has a page that gives me strength and quality of signals for all satellites it can see, whether they are fully locked or not.

On some open ground, run a suitable GPS receiver, look at the satellite quality, and then bring your dashcam up to 3m, 2m, 1m, you get the idea. This will enable you to get a qualitative measurement of whether it is indeed the dashcam, and then how well anything you do to it works.

It might be that energy is being emitted in all direction through the plastic case, in which case you're lost. It might be that most of it is conducted down the supply cable, and a suitably sized ferrite 'stopper' on the cable will improve things no end.

\$\endgroup\$
  • \$\begingroup\$ I've found that USB GPS receivers for a computer (just a round puck on a 6' USB cable) can provide a lot of detailed signal and satellite information. Consumer GPS units don't want to "clutter their interface" with all the "techno jargon", but the USB receivers will normally have a way to access all of the data they have available to them. \$\endgroup\$ – JPhi1618 Feb 11 '16 at 15:35
  • \$\begingroup\$ Not that you mention it, I do have a page on my GPS with visible satellites and bars in front of each satellite indicating how well the reception is. Thanks for the idea. \$\endgroup\$ – Dmitry Grigoryev Feb 11 '16 at 16:03
8
\$\begingroup\$

Look at "SDR dongles" - USB dongles made ostensibly for laptop reception of TV broadcasts, but repurposed for amateur radio purposes and sold for tens of dollars.

Not all of them cover the spectrum up to GPS frequencies but some do, so shop around.

Combine "SDR dongle" with "spectrum analyser" and "open source" search terms and I think you'll quickly have an outline of a solution - not a precision tool but something capable of diagnosing EMC problems and, if necessary, pre-qualifying equipment before expensive compliance testing.

Without doing any research I would suggest it's at least five orders of magnitude more sensitive (detecting microvolt signals) than your schematic of a diode-based wavemeter (detecting 0.x volt signals), so it should show most local interference even if it can't display a GPS carrier as a peak on the spectrum analyser.

A couple of minutes on Google suggest that they ARE sensitive enough to detect GPS signals presumably through the usual GPS techniques for extracting information from the noise floor. This doesn't prove much in itself but gives me some confidence that this is a viable approach.

If anyone has more specific and detailed recommendations, please add a better answer (if it isn't too close to a shopping Q&A).

\$\endgroup\$
  • \$\begingroup\$ Very interesting, though it looks harder than I expected. I think I will simply try to put a layer of alu foil on the dashcam and see if it helps, meanwhile, I'll be searching for a suitable TV tuner to buy and play with. \$\endgroup\$ – Dmitry Grigoryev Feb 11 '16 at 14:05
1
\$\begingroup\$

This question is a couple of years old, but I noticed that no one seems to have identified the most likely source of the interference -- the HDMI signal coming out of the dash cam.

https://interferencetechnology.com/hdmi-cables-emi/#

https://ieeexplore.ieee.org/document/1706346/

EMI has been a notorious problem with HDMI, especially in low-cost designs. The interference produced extends easily into and above 1GHz, as you can see in the reference material above.

I would start by assuming that the HDMI output is spewing copious amounts of EMI. If you don't need that HDMI interface, you could try opening up your unit and "terminating" the outputs with a simple RC load, or cut off the end of a good quality HDMI cable and loading down the HDMI output with a suitable load.

\$\endgroup\$
  • \$\begingroup\$ Nice find, thanks. I don't have that device anymore to try though. \$\endgroup\$ – Dmitry Grigoryev Aug 14 '18 at 7:54

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.