March 25 KiwiSDR issues....were fixed, but offline again

Got the KiwiSDR back online on 3/22. It was the TPLink TL-WR802N wifi link/adapter.... This is the 2nd one I've had die after a few months of use. The last one I had replaced under warranty. This one I may pop open and see if I can figure out what broke, I don't even get a power LED flicker. 

I replaced this one with a TL-WR902AC, which is basically an updated, dual band version. Doesn't appear to use any more power when used as a bridge on 2.4ghz only, the Kiwi and the Wifi adapter were pulling under 1A combined on 5V. 

After a week online, a big ice storm came in on March 29th. The 40m loop antenna was still holding when I left the site, but had ice built up on it. Lost power on March 30th at 7AM there, and still another day to go with ice predicted. Crossing my fingers that the antenna wire makes it through! We'll find out when the power is restored.

-Nate

3/30/25

KiwiSDR Issues March 2025

 N8BTR Crivitz KiwiSDR is down for now. Went offline suddenly just after 8PM the other night. Not sure if it's a power supply fault, or another fried Wifi Adapter, or something else. I'll find out next time I'm able to visit the site.

Either way, another nudge to get the KiwiSDR onto a wired LAN connection this year, I guess! 

-Nate 

3/6/25

A reminder to ALWAYS check components!

 On my recent QDX-M build, I was careful, inspected joints as I went, cleaned things, took my time. But once I got to the connectors, I figured I was nearly done, and I soldered them in relatively quickly. And then was surprised when the BNC plug wouldn't latch! Guess I should've checked the BNC connector BEFORE I soldered it in, this one was  REAL bear to remove with my iron! Fortunately, I had a spare from an early QDX build I screwed up. 

 

So let this be a reminder to others...ALWAYS inspect your components BEFORE soldering!

Building a QRP-Labs QDX-M for 160m

Just wanted to document what I did for a QDX-M build for 160m. I haven't seen or heard of anyone else operating a QRP-Labs transceiver on 160m yet outside of WSPR transmissions. 

      I had come up with LPF values and BPF values to try for a 160m version, but then Hans came out with the QMX+.  I already had the QDX-M unit, and the QMX firmware isn't ready for easy completely remote digital operation yet like the QDX, so I built the QDX-M for 160, and decided to NOT reinvent the wheel, but instead use the QMX+ filter values (mostly.) 

BPF: Used QMX+ values....680 pf capacitor, and 44 turns of 28 awg on a T50-2, but then removed the 680pf and installed 1200pf to bring the peak down after initial testing.

LPF: Used QMX+ capacitor values. Inductors I used T50-2 (Instead of the T37-6 on the QMX+) and reduced the turns to 30 and 27 to try and match the inductance. 

Power output @ 12V: 3.55w. Tweaking the LPF inductors didn't really affect this at all. 

BPF values: -0.5 db from peak (now)

Image rejection: -41db

 

When I initially built it, the RF Sweep looked like this: 

 

So doing some math, I removed the 680pf capacitor for the BPF, and installed a 1200pf instead. That brought the peak down close enough for me....down ~1db from the peak instead of 5db:

 

Image sweep was good:

 

Power output was much less than I anticipated based on people's QMX+ specs, getting only about 3.5w @ 12v. Thinking this is most likely due to some flaw in how I wound the RWTST, but could be something else, too. Either way, it's nearly fine as it is, as I'm using it to feed a Hardrock 50 amplifier. I ran it at 7.5V (wound for 12v) and got about 46w out. When I increased it to 9v, I had just under 2w and got 52w or so output from the amplifier.

 

 

Specs/Mods:

• Rev 5 PCB, 1.10 Firmware. 
• 28 awg wire on everything. MUCH easier on T1 and L14 that way!
• 12V RWTST using BN43 core. (Not 61 mix like the QMX+, just what it came with.)
• TN0110 finals instead of BS170s, and Wakefield heatsink.
• Amphenol USB Connector (Jeff Moore's favorite.) 
• 1N4148 diode across L14.
• PTT resistor swap. 470 ohms was too much to trigger the PTT on my MFJ1708SDR. I removed it and installed a 100 ohm 0603 resistor.
• BPF: Used QMX+ values....680 pf capacitor, and 44 turns of 28 awg on a T50-2, but then removed the 680pf and installed 1200pf to bring the peak down after initial testing.
• LPF: Used QMX+ capacitors. Inductors I used T50-2 (Instead of the T37-6 on the QMX+) and reduced the turns to 30 and 27 to try and match the inductance. 

• Power output @ 12V: 3.55w
• BPF values: -0.5 db from peak now
• Image rejection: 41db

 

Performance: I strung up a temporary full length dipole at 13 ft up at my cabin QTH, and ran FT8 for 2 nights, and a couple WSPR transmissions. The VERY low dipole did better than I thought, netting me 161 contacts in 38 states and 4 Canadian provinces with the 47 watts or so. I'll take it! Happy enough with it. Down the road, I'll install a permanent 160 antenna at that QTH, probably something vertical for DX possibilities.  (I wrote about the dipole and qso specs here: 
https://n8btr.blogspot.com/2024/12/experimenting-with-low-dipole-on-160m.html  )

 

So there you have it, a QDX-M for 160m that seems to work well. Again, power output could be better, but thinking that's a consequence of one of my actions or component choices. Receive sensitivity seemed to be very good, didn't see any better on my KiwiSDR connected to the same antenna. 

Experimenting with a low dipole on 160m

      It seems to be well known that a low dipole is a "cloud warmer," and it appears a vertical is the way to go for low band DXing, especially on 160m. But I had a hard time finding much for actual results that people achieved with low dipoles. I'd read about people trying them, and a couple being surprised how far they got on 160, thinking you'd be lucky to get 200 miles... I felt the same from what I had read about the importance of height. So I'm sharing my experience to add another reference/data point for others. 

I strung up a full size dipole for 160m with cheap 18 awg wire and a BNC to binding post connector. I only had a short 20ft RG-316 feedline, so I had to bend 1 leg of the dipole to fit in the property. One leg was about 124 feet straight, the other was mostly straight for about 50 feet, and bent about 85 degrees for the last 70+ feet. I put it up with the ends and the center about 13 feet off the ground...plenty of sag on the longer run, probably 2 ft+.  See EZNEC diagram below:

I was doing this to try out my newly built QRP-Labs QDX-M for 160m, and had it feeding Hardrock 50 amplifier. Was getting about 47w out with about 1.4w in.  

I worked 160m FT8 for 2 nights while staying at my cabin QTH where I did this experiment. Not all night by any means, but from near sundown until I went to bed, with breaks here and there, and then for an hour or so when I woke up before sunrise.

Results over those 2 nights: 161 completed contacts in 97 grids; 38 states worked as well as 4 Canadian provinces. Longest contacts were over 1700 miles to British Colombia, California and Oregon. Shortest was about 50 miles.

QSOMap of contacts:

PSKReporter shot of stations that heard me on night 1:

There was one station in Israel both nights that heard me as well, but that was the only DX station reporting me... and I heard one station in Mexico, and a couple in Spain...otherwise everything was in the US and lower Canada.

While I confirmed what we already collectively knew: a low dipole won't work on 160 for DX...it worked better than I expected on stateside FT8. On the other hand, if this were SSB, I think I would've made about 9 contacts instead of 161... Only 8 in the log had +8db reports or more in EACH direction.

So there you go... a full length, but bent dipole 0.024 wavelengths off the ground, combined with an amplifier and FT8 did OK.... Something I may keep in mind for a bit of an "extreme" POTA activation in the future. More importantly, it confirmed my QDX-M worked, and worked well. 

Some other notes: When I moved from the dipole back to my 40m Horizontal loop at 34 feet up... noise and signal levels on 160m dropped by about 12 db. And AM broadcast stations were overloading the front end on my KiwiSDR at night when fed by the large dipole...I barely see this on the 40m loop. Just interesting to me to note how much stronger receive is on lower frequencies with that longer wire!

Year 1 on HF

     I started doing some shortwave/ham listening in fall of 2023 on online SDRs. I was licensed as a Tech originally in 2005, but let it lapse...so had to test again. Took my Tech and General exams in the same sitting the end of October 2023. Got on the air a few days later, first HF contact on 11/3/23.

     I've learned a TON this past year. I've had some challenges, but also have derived a lot of enjoyment from flinging some RF watts around the world. I've enjoyed building, experimenting, reading, operating and listening.

Operations/Setup summaries for my 1st year, FT8/FT4 only, save for 1 JS8Call contact:

 I started with a 4-5w QDX that I used only on 20m, with an attic dipole and an old laptop. Added a 25w amplifier soon thereafter, but still limited to 20m most of this past year. 

I later built a portable QDX for 40-15m, and a QRP-Guys EFHW Mini Tuner/Antenna for the same. 

Ended that 1st year getting a remote station up and running. I can now operate 40m through 10m between 2 QDXs, with 40-65 watts from a Hardrock 50 amplifier (and ATU) into a 40m horizontal loop at 34 feet. Also hosting a KiwiSDR2 there. 

1st year HF Stats, all FT8/FT4, and 85% of it on 20m:

• 4617 total contacts/QSOs, 
• 1067 POTA contacts "hunted" (with the 2fers/etc included)
• 2 POTA activations with 31 contacts combined
• 104 DXCCs/countries worked, 90 LOTW confirmed. 
• 279 band slots worked, 214 LOTW confirmed. 
• Worked All States from my home shack on 20m (EN62), and 47 confirmed from the remote shack (EN55). 
• 5 Band WAS: 89 of 250 confirmed. (on 4 bands, I didn't get on 80m this year) 
• Transmitted from 6 different locations: (Home, my cabin, 2 AirBnBs in Illinois, and the 2 POTA activations...one in Missouri and one in Milwaukee where I live.) 

 

Notable things set up/built, etc:
Built 3 QRP-Labs QDXs, one for 80-20m, one for 20-10m, and another I made for 40-15m.  

Bought and set up a KiwiSDR2.

Antennas: Set up a 40m loop antenna at 34 feet at my cabin / 2nd QTH. Set up a 20m speaker wire dipole in my attic at my main home. Built a linked EFHW for 40-15 for the QRP Guys Mini-Tuner. Mostly built a Pacific Antenna 40/20 trapped dipole... but didn't deploy that yet. It's slated to replace my 20m attic dipole. 

Also built: 

• Emtech ZM-2 Tuner
• 10th Anniversary Hardrock 50 amplifier with prebuilt ATU add-on
• DJ0ABR TRX Selector Board (SMD soldering!)
• Relay/control box with Devantech ds378 relay board, voltage regulators, etc. Basic Node Red flow work, too.
• QRP Guys EFHW Mini-Tuner for 40-15m
• QRP Guys QRP Power/SWR Meter/Dummy Load
• Pacific Antenna/Kits 0-11db attenuator
• K9DP BCI Filter
• QRP-Labs 20w Dummy Load

"Goals" / Thoughts for year 2: 

• Continue to operate/improve the remote station now that it's up and running on 2 QDXs. (Filling out more band slots now that I can operate routinely outside of the 20m band!).. hoping to get to, and beyond 100 DXCCs confirmed. 
• Set up another antenna, switch, and QDX to operate on 80 and 160m. (Need to build a 160m QDX). 
• Based on performance of existing antennas....evaluate my next antenna move: 80m loop? 160m loop? Leave the 40m loop and a separate 80/160m antenna? 
• Evaluate my next transceiver move: A QMX+ for SSB and 6m coverage? Hermes Lite 2? Save for a Flex, now that my cabin has fiber internet, and thus not speed limited?
• Do a bit more POTA/portable/travel operating... another couple activations maybe?

KiwiSDR antenna issues solved

    I pulled down the 40m loop entirely to inspect it after sweeping SWR and finding nothing there. Found nothing physically wrong, really. I added some electrical tape where water was getting into the PL259 threads at the balun, and soldered a connection that was just crimped. Checked the 80 feet of coax with a NanoVNA, and got 1.95 db of loss at 50 mhz, and about 1db at 15 mhz. Normal for RG-8X with a GDT in it. 

     I found a couple issues in the shack: A PL259 jumper cable had a loose ground connection to the shield, and a right angle BNC to PL259 jumper was shorting out or something depending on position. I could turn the interference/noise on and off by moving the cables around. Hard to say if it was one or both, as they both plugged into an MFJ SDR switch, and thus moving one kinda moved the other. I replaced both cables, and signals were back to as they used to be. 10m WSPR SNRs improved by 25+ db, back to where they were before 10/13. 

I've vowed to use more flexible (smaller cable) jumpers where I can in the future, as the thicker ones are wrenching around light things they're attached to. So more 6-12" RG-174 jumpers instead of RG-58 or RG8-X.