Tag Archives: noise

miniVNA and Choke Measurements

I recently bought a miniVNA Pro analyzer:

minivna_pro

This is a pretty cool little device that goes beyond just analyzing antennas. One of the first tasks I set this to was measuring some ferrite chokes for comparison. The miniVNA makes quick work of this in transmission mode and in saving the data in Excel format. Here are choking impedance plots for three ferrite based chokes:

chokes

In blue is 9 turns of RG-8X around two FT240-43 cores. This has the best choking performance. The orange plot is a 1:1 choke balun made from an ebay kit. The grey plot is an MFJ-915 line isolator.

Advertisements

2el Vertical Beam Antenna WSPR Results

I ran the 2el vertical beam antenna on 20m WSPR for most of the last week. This time I looked at the WSPR challenge data as compared to N5CEY’s station about 15 miles away. During the week we both made the WSPR challenge scoreboard. On some days I had a higher unique count while on others N5CEY did. The WSPR Challenge site provides a list of all the stations heard in a 24 hour period along with their counts. This can be downloaded and sorted to see which stations had the highest counts. The data for a specific station can then be downloaded over the time period. Here is the count comparison:

wspr challenge counts

Here is a plot of the reported SNR vs time to K8JBV in Ohio from both my station (red) and N5CEY (blue):

WSPR comp

This chart is fairly typical and shows that N5CEY has about a 6dB higher SNR at his station vs. mine. I did note at least two stations, one in Georgia and the other in Florida, that had SNR roughly equal at the two receiving stations. This correlates well with the direction I have the vertical beam pointed. Likewise stations on the backside of the beam have a  few dB more separation.

I am pretty happy with the performance of the vertical beam antenna overall. What I would like to do next is operate portable from a more rural location and measure and differences in SNR on WSPR.

Measurement and Comparison of Common Mode Chokes

I have several common mode chokes in the shack today that I wanted to measure and compare against each other. This seemed like a good application for the low cost VNA I bought a couple of months ago. The setup is shown on the N4SPP website. Only the shield oif the chokes is connected to the center pin of each VNA input. The sweep is done in Transmission line mode and the result is Transmistion loss and phase. This was run after a calibration sweep using a short. I has some concerns over the overall accuracy but the results relative to each other is a good start for comparisons. The chokes I measure are two ugly baluns:

IMG_1487

12T & 14T of RG-8X on an FT240-43 Toroid:

IMG_1486

A 1:1 balun kit I ordered and made off of Ebay (after being outdoors pretty ugly in its own right):

.IMG_1489

and finally the guts of a MFJ-915 line isolator. I say the guts because the isolator originally was encased in PVC pipe but the connectors were off center and did not allow installation of the mating connector. I cut this off and wrapped the connectors in silicone tape.

IMG_1488

This graph shows the resulting data on sweeps from 3Mhz to 15Mhz:

2018-03-19_15-05-13

Overall, looking for damping in the range of 25-30 dB. At first glance all of these chokes all look about the same with the exception of the MFJ line isolator which although relatively flat has readings over 20 dB. G3TXQ has a great write up on the choking resistance. He notes that an issue with the “ugly” balun is that their impedance is almost entirely reactive with the exception of a small band of frequencies around resonance.

“Reactive chokes have the disadvantage that they can “resonate” with a CM impedance path that is also reactive but of opposite sign – in some cases actually increasing the CM current flow rather than choking it.”

I removed the ugly balun and replaced it with the 14T Toroid Choke at the feedline and added a 12T Toroid choke at the IC-7300 antenna input with a 1 S unit drop in the noise floor noted. I will replaced these with mix 31 toroids of the same size which should give a flatter, broader response below 15 Mhz. I also have a 1:1 balun on order from Balun Designs for use at the feedpoint. I’ll take some measurmeents on these and post and they become available.

Lowering the HF Noise Floor

Previous posts have reported on WSPR performance with my vertical antennas especially compared to data collected by a nearby Ham (Cliff – N5CEY) with a similar antenna setup. The metric is how well a given antenna receives unique spots over a 24 hour period as documented by the WSPR Challenge site. I have always seemed to lag behind Cliff’s receive performance in spots and in DX received. I have asked Cliff about the specifics of his setup and came to the conclusion that he has a significantly lower noise floor at his QTH in the country versus mine in the city. Cliff’s QTH is about 12.5 miles north of mine in a rural area. He is closer to the coast and likely has much better soil conditions as well. His antennas on WSPR are all 1/4 wave verticals on 40m, 30m and 20m. I have an “ugly” balun installed at the feed point of my vertical and about 30 feet of RG-8X to the shack entrance. There are about 8 loops of coax at the shack entrance. Inside the shack are various patch cables connecting switches, antenna tuners, SWR meters and the Timewave ANC-4 to the IC-7300. Lots of nearby noise sources including PC’s, cable modems, wifi routers etc.

So the task before me is how to lower the noise floor at my QTH to bring in weaker signals. I have done quite a bit of research and finally came upon some discussion of the effects of Common Mode currents on noise floor. My next post will detail some measurements on various common mode current chokes but for now suffice it to say that changing my common mode choke arrangement has had a positive effect on the noise floor. Here is what I did:

  1. Added a CMC consisting of 12 turns of RG-8X around a FT240-43 toroid core right at the IC-7300 antenna connection. This resulted in a 1 S (6dB) unit reduction in the received noise!
  2. Replaced the ugly balun with a choke as above. No significant change noted in noise floor.
  3. Replaced 30 feet of RG-8x feed line to the shack with 30 ft of RG-213 coax. No significant change in the noise floor was noted. this change should help in reducing losses though.

Adding the choke to the antenna input made a big improvement in lowering the noise floor. I adjusted the Timewave ANC-4 to further try and negate any remaining noise.

So the results are as follows after 20m WSPR was run for 24 hours:

wspr1wspr2

Pulled ahead of N5CEY by 8 unique spots and made the DX list still lagging a bit behind Cliff. Here are my spots during this period (right image is Cliff’s and the one on the left is mine):

2018-03-18 (1)2018-03-18

So all in all a significant improvement by addressing Common Mode Currents. More on the Chokes in an up comming post.

 

The S9V31 is Working…but How Well?

I have been trying to collect some solid data on the performance of the S9V31 antenna configured for multi-band operation. It is currently configured to be non-resonant on any band and I go through an antenna tuner in the shack. WSPR results have been lackluster and possibly a little worse than when run as a resonant 40m antenna. FT-8 spots on pskreporter.com on the other had have been quite good on early morning 40m. I am being heard all along the western Pacific rim consistently.

I have noticed quite a few stronger European stations in the late afternoon on 20m and have made several SSB contacts. I actually closed four all time new ones this week: St Eustatius on 12m, Republic of Congo & Easter Island on 17m, and Revillagigedo Island on 20m. Interesting that 12m and 10m have had brief openings in the late afternoon. Band conditions have been pretty bad with long periods of zero sunspot numbers.

I am still working on what to do to lower the noise floor at the shack. I feel that is the key to improve station performance and should be readily visible when running WSPR. More on that coming up….

 

A Quick Radial Experiment

I added about 12 feet to two of the 40m quarter wave radials, trimmed the ends so there is bare copper and dropped then into the resaca in the backyard. For those not familiar with the word “resaca”, it is a type of oxbow lake which in my case was a channel of the Rio Grande River that was cut off from the river and forms no inlet or outlet. The water is fresh but somewhat brackish.

The result appears to be minor. SWR dropped to 1:1 but not sure if it is because of the longer wires or the contact with water.

20m Vertical Beam Now Pointing to Japan

I moved the driven element of the 20m vertical beam to a bearing of about 330 degrees today. it took only about five minutes to drive in the pvc mounting pipe And rearrange the ground radials. The driven element is now further away from the A/C noise so that is now tolerable.

Early morning in Japan, signals started coming in and this time the difference between the vertical and the beam are significant. The S-meter indicated about a 2 unit difference. I did not close a QSO on phone but made a JT-65 to Japan.

Tomorrow I will collect some WSPR data and share a comparison method I am trying out.

 

My Never Ending Battle with Noise

I had to replace my central air conditioning about a month ago. It is a high efficiency type. The 20m vertical beam I have been experimenting with has the driven element about 25 feet from the outside compressor unit with the air handler about the same distance away just inside the house.

Whenever the fan is on (currently in heater mode) I get these noise spikes evenly spaced every 18khz across the 20m band that is being picked up by the beam. The vertical which is much further away does not pick this noise up on 20m but does see it on 18m. The Timewave does not cancel it out as the sense antenna does not hear the noise in its position inside the shack.

I’ll be moving the driven element of the beam to point towards Asia tomorrow which should improve things as it is further away from the A/C. Turns out that high efficiency A/C units these days tend to use PWM circuits for the fan control to optimize efficiency. This plays hob with RF above 20m.

Sigh…

ARRl Sweepstakes 2017 – Roundup

The 2017 ARRL Sweepstakes was this weekend. I had some personal issues to deal with on Saturday and did not actually start working the contest until Saturday evening, 5 hours after the start of the contest. At this time of day 20m and up were closed down at the QTH. I started by working on 40m and spent most of my time there. The 40m 1/4 wave vertical worked great.I made contacts easily and the exchange was copied 100% about 95% of the time. I tried 80m on both the vertical and the DXtreme end fed and while both tuned up initially, there were problems when it came time to actually make contacts. The auto-tuner would start trying to re-tune every time I went to transmit. I only made one contact on 80m to North Texas. I worked Saturday until about 11:30 pm.

I ran a few more contacts on 40m early on Sunday morning before I left for church service. I picked up again in the early afternoon and started working on 20m and 15m. 20m was very busy and there was considerable QRM. 15m was open but there were significantly fewer stations. I struggled to make contacts initially but after about an hours they started to come in easily. I suspect that this was due to propagation changes. I quit around 5:30pm after making a few final contacts on 40m.

All in all, I worked the contest for only 6 hours and 33 minutes, made 124 QSO’s across 63 of the 83 available sections. I worked enough “rare” states that I completed my ARRL WAS for the phone endorsement using only LoTW confirmations.

What worked:

  1. Very pleased with the 1/4 wave vertical, with the IC-7300 and received several complements on my audio quality.
  2.  The Timewave ANC-4 has become easier and easier to setup and use. It is very valuable when noise conditions changes during the day. I am finding it to be effective at reducing shack noise.
  3. N1MM logger+ is very very handy for contest especially with the spectrum display on the PC with the IC-7300

What I need to improve:

  1. I don’t have a capability to work 80m and below effectively. This is a problem that I need to address as there were rich contact opportunities on 80m during the evening hours. I thought I could use the DXtreme for this purpose as I did in 2015 but I suspect there is something with that antenna that has changed. It may be time to bring it down.
  2. My foot gets tired using the foot switch. I switched back to the hand switch occasionally but I need a better switch for this purpose.
  3. Need to work on adjusting the receiver when there are many stations close in causing QRM.

I don’t really compete in contests for the prizes or fame. I work contests to test my setup over various bands and look for areas that I can improve. I think I was successful in that this year. That being said I was only 20 sections from a clean sweep mug so who knows? maybe next year?

 

 

 

Initial impressions of the Timewave ANC-4 Antenna Noise Canceller

My used Timewave ANC-4 arrived yesterday from Ebay. It lacked a power connector which fortunately I had in my junk box. I wired the cable up and added some power poles. It came with an antenna that consists of about 8 feet of wire connected to the center pin of an RCA plug. I set this up roughly vertically inside the shack. The unit must be grounded to work effectively with this wire antenna. I fiddled with it awhile but did not have much luck in reducing noise initially. Per the manual, the first thing to try was ensure that the sense antenna was picking up the noise at about the same amplitude as the main antenna. The receiver is setup to an unused frequency and the noise level measured on the S-meter. The phase and noise gain controls are fully CCW. The  main antenna was disconnected and the noise gain adjusted until it matches the S-meter reading  measured previously. The main antenna is then reconnected and the phase control is adjusted until a null is found.

This took awhile to figure out but with the IC-7300 it is easy to see the results on the waterfall. As you get close to a null the background on the waterfall will darken and signals will become more distinct. I estimate that the change is about a 2 S unit reduction in noise. I have run the unit on 20m WSPR and over a 1 hour period had 42 unique spots compared to N5CEY’s 21 unique spots.

So initial results look promising. I will need to work more with this and collect some data on to its effectiveness.