EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: D3f1ant on September 01, 2015, 11:15:19 pm
-
Anybody have any experience or recommendations for some basic antennas?
I'm leaning towards this kind of thing, simple single antenna solutions.
http://www.aaronia.com/products/antennas/BicoLOG-30100/ (http://www.aaronia.com/products/antennas/BicoLOG-30100/)
http://www.aaronia.com/products/antennas/BicoLOG-30100-X/ (http://www.aaronia.com/products/antennas/BicoLOG-30100-X/)
Do I need/want the pre-amp with a Rigol 815?
The goal is to be able to take some outdoor measurements at distance, its pre compliance so absolute accuracy not as important as a getting a general picture of likely hood of passing formal pre compliance scanning. Anything cheaper worth considering, something from Asia?
-
I went through fcc b last year on a digital board (1ghz cpu, ddr3, ddr2, 4 hdmi, 8 usb 2.0, various dc : dc switching supplies etc). What I found was that a small beehive em loop plugged into my dsa815 was enough to sniff all the various 'openings' in the case of the digital board and to check leakage at the various connectors.
I used a DC LISN to sniff both the 'backwards' and 'forwards' noise from the AC adapter and also coming back out of our box. Depending on the power source of your product you'll want to precompliance test for conducted emissions.
Everything I observed with the simplistic tools above matched what was observed at the compliance testing lab. After replacing the AC adapter (not our product) with a 'better' unit we passed the full test (took some 5+ hours to run the various tests).
So, I'd suggest it is possible to go a long way in testing a product for gross issues (really depends on what that product is and you haven't provided any info here...) with a fairly inexpensive set of 'accessories'.
We did have considerable input on various PCB layout 'tricks' from an RF/Microwave engineer. That certainly must have helped since our digital board was very quiet and all the issues were external (AC adapter and sourcing quality HDMI cables etc).
The actual compliance testing used various antenna assemblies that could raised/lowered/moved forward and backward and the DUT was on a rotating table. The whole compliance test was automated and quite a 'dance' to watch via CCTV from outside the chamber.
cheers,
george.
edit: fix typos
-
This web page has some DIY antennas. Unless you have a proper chamber the measurements on a desk are going to produce ballpark figures at best. http://www.compliance-club.com/archive/keitharmstrong/emc_testing1.html (http://www.compliance-club.com/archive/keitharmstrong/emc_testing1.html)
The best way is to do a pre-compliance quick scan at an EMC test house. That will give you a graph with potential problem frequencies and their level. With the simple antennas you can scan your PCB to find the problem areas and try various solutions to get the emissions down enough to pass.
-
Thats what we have done in the past, ie get a pre scan done and then investigate any problem area's with near field probes. We also already have DC and Mains LISN's (Tekbox) for conducted. The Antenna idea is more for scans of prototypes at the stage we wouldn't want the expense (or time delay) of a formal pre compliance scan on, especially we have a tool that should be able to do it. Thought about a TEM cell but I think that would be essentially useless without a reference point of a proper calibrated scan. The Aaronia antennas seem to have calibrated data with them, and they look pretty :-DD
In New Zealand, its still easy enough to find places 1/2 an hour drive away completely devoid of RF :) so I was interested in being able to do scans more whimsically and very early on in the design cycle, ie prototype or even proof of concept boards to get a feel for them. The products vary wildly, but generally wouldn't have oscillators higher than 100Mhz, hence the 1GHz limits are fine.
Thanks for the responses.
-
The big problem with small 'sniffing' antennae is that they're used close-up. This is fine if you're trying to diagnose the source of a spur you saw in the test cell, but you can't extrapolate from an overall near-field measurement to a far-field one. Fortunately, anything you can suppress in the near-field will also be gone in the far-field, but if you use that to get compliance you'll work unnecessarily hard, removing emissions that really aren't a problem.
Wanting to get some roughly calibrated results in my own workshop to compare with the test house I picked up a secondhand biconical antenna. They're not easy to find, though, at a low price. And huge compared with those wonders from aaronia. Then I discovered it's actually surprisingly easy to calibrate your home made antennae - you don't need a gold standard to compare against. You just need two identical (or three non-identical) antennae and something called the three-antennae method to determine the frequency response of your antennae.
-
I have used an Aaronia EMC kit before, including their antennas for this kind of work and had great success.
-
For small boards we are using this temcell for pre-compliance measurements:
http://www.tekbox.net/test-equipment/tbtc1-tem-cell (http://www.tekbox.net/test-equipment/tbtc1-tem-cell)
Despite the limitations, it produces quiet good results.
-
In New Zealand, its still easy enough to find places 1/2 an hour drive away completely devoid of RF :)
Even no FM radio? A cable of a few meters is a perfect antenna for FM radio. Anyway, if you can make a setup in a remote area you might be better off with a (big) wideband antenna.
-
30MHz to 1GHz? :-DD :-DD :-DD Yeah, with some pretty serious calibration (and lack of SNR), maybe! What a ripoff... 300MHz at best!
As long as you aren't concerned with < 300MHz stuff, or have another antenna to fill in that region (one of those cagey conical dipoles, 1-2m long, would be fine), it would still be handy, though. And the usual E/H probes, for near field studies.
Don't forget polarization when using the antennas, of course. :)
For pre-compliance, I'd also suggest that all of these can be hand made. The traditional bowtie or conical shape is nothing special, and should calibrate the same either way (if you wish to get it done).
Tim
-
Yeah the 'cheap' Aaronia antennas seem to have a 'bit' of drop off in the low end, not sure -39db is useful :P. I can't find any sort of reasonably priced antenna or antenna set, there some old bipoles on ebay but they still fairly spendy. Might have to look at DIY antenna approach during quiet periods at work, never built an antenna before so that could be interesting anyway.
Where are the low budget antenna's (ie china #1 brand/clones) to go with the low buget Rigol?
-
Try HAM antennas instead. Wideband is good for receiving many stations at once (or with many receivers from a common antenna).
Tim
-
you can't extrapolate from an overall near-field measurement to a far-field one.
Sure you can. It's done every day. It's just that spatial Fourier transforms tend to be out of the reach of most hobbyists, and calibration can be difficult.
30MHz to 1GHz? :-DD :-DD :-DD Yeah, with some pretty serious calibration (and lack of SNR), maybe! What a ripoff... 300MHz at best!
As long as you aren't concerned with < 300MHz stuff, or have another antenna to fill in that region (one of those cagey conical dipoles, 1-2m long, would be fine), it would still be handy, though. And the usual E/H probes, for near field studies.
ETS Lindgren makes a "biconolog" antenna that is essentially a hybrid between a biconical and a log-periodic. They have decent response from 30 MHz to ~5 GHz and are about 1m long. Above 5 GHz, a pair of ridged horns will get you up to 40 GHz, which is really the highest frequency that any of the EMI standards require. Of course, none of those antennas are in any way cheap...
It's not uncommon at all to see 70+ dB amplifiers in an EMC lab. As long as your noise floor is at least 6dB below the limit line, you're fine.