HFU HF Underground
Technical Topics => Equipment => Topic started by: bandarr3000 on December 13, 2017, 0202 UTC
-
I have an SDRPlay RSP2 which has the Hi-Z port (with 1k ohm input impedance). I use 40 feet of 50 ohm coax to a 1:9 balun and then two wires from the balun going into the Hi-Z port, which I figure gives me 450 ohms of impedance at the receiver (I find that this setup reduces the noise when compared to straight coax to port A on the RSP2). With the coax and the balun, what impedance am I really matching my antenna to? 50 ohms? 450 ohms? Something else?
-
Please provide a little more info, specifically what type of antenna and how does the 50 ohm coax attach to the antenna.
FWIW, my RSP1 has only the single 50 ohm SMA input so, my wire antenna (450 ohm) connects to a 9:1 balun to RG-58 (50 ohm) coax to the RSP1.
This set up works fine on HF but seems to pick up lots of noise on LW and MW, however, the low pass filter in the RSP1 is 12MHz whereas the low pass filter in the new RSP1A is 2 MHz.
In other words, they seem to have realized the need for more filtering.
So, my inclination is not to blame the antenna so much as perhaps the relative lack of adequate filtering in the first version SDR.
What freqs are you most interested in receiving and what is the freq of the low pass filter in your RSP2?
-
I'd never chance the rfi from a two wire input setup to feed my rsp2. Straight coax all the way, with ferrets at each end of the coax. I have pretty sever rfi from the pc and networking gear so wouldn't even consider a balanced input to an sdr that is two feet from the back of the pc.
-
Thanks for the responses. To add more info:
1. The antenna is a horizontal "delta" loop (not a perfect square or circle) about 170 feet long hooked up to one of Chris Smolinski's Squid transformers. I listen almost exclusively to HF.
2. I have a 2.6 Mhz high-pass filter on the coax before the 1:9 balun. This attenuates the AM band pretty well I think. I lose MW and LW, but I'm in the middle of the country and don't get any LW anyway.
3. On the two wires between the 1:9 balun and the Hi-Z port I have a large type 43 ferrite. Wrapped about 8-10 turns if I remember correctly.
4. I also have ferrites on the coax just before the high-pass filter, and on the USB cable just before the RSP2.
5. I seem to get a lot of noise when I plug the coax directly into port A. Also, I don't have a ground when plugged into port A. I use the ground on the Hi-Z port when I connect that way.
6. I live on the outskirts of a large city.
-
Lots of taps on the Squid, so, let's assume that by experimenting you have found the best match to your horizontal loop. (The idea here is that when impedances are matched there will be maximum transfer of signal).
From the Squid, run a good quality coax to your RX location. "Good" means not only low loss but also well shielded. Some coax has very minimal outer braid shielding.
Your noise could be from nearby radio transmitters or locally generated electrical/electronic QRM or both. If you have noise pickup problems, consider using a common mode filter (choke) as close to the RSP2 RX as possible. Lots of info on the internet on how to build a common mode filter or you can purchase one already built.
Leave the ferrite on the USB cable or get a better quality (shielding again) USB cable. This cable can pickup lots of noise and transfer it to the RX.
Looked at the specs for your RSP2 and like my RSP1 it has a 12 MHz low pass filter. So, especially below 12 MHz you need some help. Have found here that an antenna tuner at the RX input helps considerably in peaking the desired signal. Pulling it out of the noise so to speak. Even something as simple as an MFJ 956 works wonders.
Don't know where on the outskirts of GR your QTH is located, but if it is near any of the half dozen or so AM broadcast stations in your area, you may have to build a notch filter(s) (also known as a wave trap) specific to the frequency of the offending station(s).
If the antenna is bare stranded wire, in time the wire will oxidize between the strands and create high resistance and a noisy situation. Copper oxide is a good rectifier... think mixing products.
Use bare or insulated solid wire or insulated stranded wire.
Just a few thoughts that hopefully may be of some help.
-
I think I've found the best match to my loop antenna with the Squid, but the reason for this post is to figure out how to calculate the impedance so that I know I have the best match. Watching the signals on the SDR helps, but is not completely reliable.
For coax, I have RG8x and RG58 with a bulkhead in between. I'm strongly considering an antenna tuner. As for the loop, it is 16 gauge insulated
stranded THHN wire.
Thanks for your help. I truly appreciate it.
-
By calculation, (f=1005/ft) your 170' full wave horizontal delta loop is resonant at about 6MHz and impedance is theoretically about 100 ohms at resonance. Actual impedance may vary from maybe something like 50 ohms to over 200 ohms based on various factors (mostly height above ground). An antenna analyzer would be the best way to determine resonant frequency and actual impedance at any given frequency. However, a good match can be obtained by experimentation. Tune in a steady continuous signal such as WWV at 5000 or CHU at 7850 (daytime would be best as there is less fading in and out) and try different taps on the Squid while observing the S-meter or better yet the SNR (signal to noise ratio) on your SDR. There are enough taps available so you should find something that will provide a close match. In other words, you don't really have to know the exact value of impedance because you can get a close enough match by experimentation and, also, the impedance will change with frequency. My suggestion would be to match the antenna as best you can with the Squid near its resonant frequency and then use an antenna tuner to match to other frequencies.
For example, with my RSP1 is connected to a random wire antenna about 50' long through a matching unun. Tuning in a weak signal on the AM broadcast band, the SDR displays a SNR of 3db. Now switching the antenna tuner (MFJ 956) in line and peaking the signal results in a SNR of 15db.
Let us know how things work out for you.
-
OK. Thanks. Makes me wonder why I find the strongest signal for 6480 kHz at the 703 ohm combination on the Squid. Wonder if it's the 1:9 balun between the 50 ohm coax and the Hi-Z port (1000 ohm impedance port) on the RSP2. Using the Squid combinations near the resonant frequency for the antenna don't get as good of a result as the 703 ohm combination. Neither does anything higher. 703 ohms seems to be the sweet spot.
-
Wonder if it's the 1:9 balun between the 50 ohm coax and the Hi-Z port (1000 ohm impedance port) on the RSP2.
Probably.
You say that you are using the high Z port in an attempt to reduce noise. You may be reducing the noise but you are also probably reducing signal as well.
Try to eliminate the balun and go directly into the SMA 50 ohm port. See what tap on the Squid works best.
You mentioned RG-58 coax. Some RG-58 is real junk, inadequate shielding. Use a good quality name brand coax. Your feedline is a likely culprit when picking up noise.
-
Ok so if I'm reading you right, you're feeding the loop to the squidlinski, then to coax, then to a hi z input?
If so, you need another 9:1 mosquidski to go from the coax to the hi z balanced input. Either that or feed the coax to either A or B inputs. The A input has a lovely ambc notch that can be enabled at will in your sdr application, so you don't have to remove the ambc filter by hand when not desired. The tiny coax connectors used on these devices isn't really designed for lots of connect/disconnect cycles.
-
Kinda late to the discussion, but as a RSP2pro owner using the High Z port, I might be able to offer some insight. Maybe not, but hey, better too much information than not enough.
I have a pair of homebrewed loops that, for the time being, are the only antennas I use with my RSP2pro. Both have RG-58 feedlines, 50 Ohms with BNC connectors (necessitating the use of BNC to SMA adapters). I use a 1:9 balun when connecting to the High Z port. I have a BNC-to-bare-wire adapter as well, but I've had almost no discernable reception when I used that to connect my loops to the High Z, so it seems that the balun is necessary when using that port. I don't know what using the Squid would do to any of this, but I bet one of the combinations would give you a good impedance match without the need for the 1:9 balun. But just in case...
The balun:
https://www.amazon.com/NooElec-Balun-One-Nine-Applications/dp/B00R09WHT6
SDRplay info sheet about using the High Z port (PDF):
https://www.sdrplay.com/wp-content/uploads/2016/12/161201HighZPortGuidev2.pdf
I hope this helps. Good luck.
-
Thanks for all the help. It seems I have other fish to fry, as I have MHz-wide RFI to deal with. Ever since the cold snap started.
-
RFI/EMI... it's everywhere. :(
If it started with the cold WX, check for a controller or such on a heating appliance. A portable receiver can be useful as a sort of direction finder to track something like this down.
-
I think I've nailed the source of the RFI. I took SNR readings right in the middle of the noise every minute for about 24 hours (using the handy feature on SDRuno). When I was away from home for the day and had the furnace turned down, the RFI was pretty much non-existent. When the temperature was turned back up, the RFI came back in a big way. I have a variable-speed DC motor on my furnace, which I'm about 95% sure is the culprit.
-
Good troubleshooting, thanks for the update.
rgg