Apr 10

The Poor Man’s SDR

This has been a long awaited post for a few people in the community who have asked me about this project. I dub it “The Poor man’s SDR“. This post will simply serve as a summary. Your mileage may vary.

The goal was to enable an already feature-rich radio with SDR receive capabilities near that of a Flex Radio or K3 with LP Pan. There has been much debate within the TS-480 Yahoo forum surrounding which IF to tap, 10Mhz vs 70Mhz. (Ask me why I no longer participate in said forums…)

The 1st IF (70Mhz) provides much more spectrum to feed into PowerSDR although required specialized hardware to even tune the 70Mhz IF. If you already have made a capital expense into a Perseus or like radio this wouldn’t be an issue. The 2nd IF is downstream from the roofing filter which limits width to roughly 20khz. I know what some of you are thinking.. “This is useless, you don’t know what you’re doing. Go die in a car fire!” I beg to differ. (Ask me about forums again…)

If you need 96k or 192k of SDR bandwidth, buy a Flex 3000 or 5000 like I ultimately did. They rock! But if you want to have some fun with what you already have, read on.

Here’s what you’ll need:

A Kenwood TS-480(SAT or HX)
A Z10000 buffer amp from Clifton Labs. I bought my buffer assembled with a 4db gain (i winged the spec). It also came with RG-178 teflon coax with connector (mini sma) already installed and an mini SMA to BNC to connect to a softrock receiver.
A Softrock receiver from KB9YIG. I’m utilizing a Softrock RX Ensemble II. This was donated from NB3R, whom ordered his pre-built
A good sound card. See this link for more info.
A good soldering iron, an anti-static wristband, some electrical tape, a Phillips screwdriver, a steady hand and enough guts to disassemble your fully functional TS-480. A stiff drink may assist with the latter.

WARNING!!! – I am not responsible if you wreck your radio. You assume full responsibility for your actions. This WILL void your warrantee. You have been warned.
Now drink because I’m writing this based on my memory of doing this project almost a year ago.

When you open up your radio (the side with the lit logo), you will be faced with the logic board. This board is where the filter board and voice keyer plugs in. Disconnect the grey coax jumpers (these are the IF lines), remove the grey cover where the voice keyer goes. Also remove the filter board.


Remove the ribbon cable card-edge-type connectors from the right side of the board. Be especially careful here. If you break the plastic clip which locks the card edge in the slot, the ribbon will not stay firmly connected in place and the radio will either not work or behave erratically. Most damage doing this mod is likely to occur here. Don’t use pimp/goon hand strength. Remove the remaining wiring to this board.

Remove the remaining screws holding the board in place and remove the board from the chassis. Turn it upside down. We will be working on the underside. What was once on the right will now be on the left (for the spatially challenged).

Read this part carefully.
Cut the teflon coax in half, solder one end of the cut RG-174 to the input of the buffer, the other to the solder side of the board where the 2nd IF coax socket is located (circled in RED). Center conductor to the small solder point (top), shield to the larger (bottom).


Route buffer so the coax to the IF tap passes to the other side of the board where the filter board would go. Route the other RG-178 (with connector) through one of the square vents on the front of the radio (with the mini-sma connector on the outside) and solder it to the output of the Z10000 buffer board.

We now need to provide power to the Z10000 buffer. I accomplished this by piggy backing power from the PWR1 (HX model) side of the radio. The reason for doing this is to provide power to the buffer even if it is in RX Only mode (HX model) with one leg of power connected.

Carefully turn over the radio and remove the panels to expose the RF deck. Locate the radio’s power leads (large red and black wires). Solder some power wire (light gauge) to the power input of the buffer and route it through the side of the chassis to reach the other (top) side of the radio. Solder the buffer power leads to the power leads of the radio. See photos.


Now that the wiring has been completed, re-assemble the bottom of the radio, turn it over and reinstall the logic board. Reconnect ribbon cables (carefully) and all other electrical connections. Wrap the buffer board in electrical tape or some other non-conductive material. Re-install the filter board, the plate which overs the voice keyer and all external covers. If you have screws left over, you’re doing it wrong.

Power up the rig. If it works, you did it right! Congrats on not destroying your radio!

Turn off the TS480 and connect the Softrock antenna input to the mini-sma connector you’ve just installed in the 480. Power the beast back on.

Tune your Softrock receiver to 10.550 which I believe is the 2nd IF frequency of the TS-480 (going on memory of a year ago). The audio output should be wide frequency audio with very high pitched sound and should change as you roll the VFO knob.

Download and Install PowerSDR-IF from WU2X. Follow through the steps he’s indicated in the “Getting Started” section. With some success, you should see similar results as below.


PROJECT NOTES
Buffer Function – The purpose of the buffer board is to protect the IF stage of the radio. It also prevents output sag in the IF path which weakens the receive capabilities of the radio. This project can be completed without the buffer, but based on the information I’ve just stated, it was well worth the $44.

SDR Bandwidth – Going back to the 1st vs 2nd IF debate, I’ve had excellent results with the approximate 20khz bandwidth provided by the 2nd IF. With applications like CW Skimmer, I was able to “hear” CW QSO’s 10khz above and below my selected center frequency. Eg, Tuning to 7.010Mhz would provide decoding of 7.000 to 7.020. The waterfall in Skimmer was much clearer with more successful decoding by utilizing IF tuning over typical 3khz audio output. With this setup I was able to dramatically increase my CW QSO rate in a contest only weeks after I completed this project.

Buffer Gain – I stated that I “winged it” when I selected 4db of gain on the buffer when I ordered it. This was based on documentation interfacing the Kenwood TS-2000 to the LPPan. I felt they may be similar.

Panadapter Anomalies – A noticeable null in signal is present in SSB and CW modes on the panadapter at the center frequency. This is not audibly present. A workaround I employed was tuning my IF receive frequency down by 5khz (10.500) on the Softrock and adding offset to PowerSDR to line everything back up. This may be caused by the Softrock. Need further testing to prove out.

Mirrored Waveforms – You may see mirrored waveforms surrounding the center frequency. I believe this to be caused by delay in phase between the I and Q outputs of the Softrock. This should be resolved by utilizing a higher quality IF receiver such as an LP Pan or like hardware.

Nov 28

New beam

A month ago during the 2011 CQ Worldwide Phone contest I lost my roof-top hexbeam to wind. Areas around me would not only suffer damage from the wind but also record snowfalls. I got off lucky, altough my hexbeam did not. Also to note, this was a homebrew antenna and not one commercially made. The hex is an excellent design and definately gets the job done, so no complaints here.

With that out of the way, I was on a search for some aluminum and recently encountered a Hy-Gain Explorer 14 (EXP-14) that was in the attic of a local ham. I knew nothing of this Hy-Gain model, but the specs seemed to be right in line of what I was looking to accomplish. The price was right and it was in excellent condition. In speaking with the seller, the beam used to belong to a local club who unfortunately lost access to put up towers for field day and this beam was one of the casualties.

The Hy-Gain EXP-14 has a unique feed point (as they even state in the manual) with parasitic driven elements on 10m. It weighs approximately 45lbs, 14ft boom with a 17ft turning radius, perfect for my rooftop mount.

I had to beam built, tuned and installed in what seemed to be record time, only taking 3 1/2 hours. Part of that time included running out to Home Depot for a 12″ mast (above the rotor) and prepping the coax with ring terminals.

Once the beam was up, I could hear a drastic difference compared to my temporary bob-tail curtain on 20m. The first stations I heard were from India and Oman. Keep in mind, all of this was going on during the CQ Worldwide CW contest that I unfortunately could not commit a serious amount of time in due to the holiday, but ended up being a good time to test the new antenna and tweak the station.

The antenna is only sitting about 25ft off the ground, 6ft from rooftop. I was concerned with takeoff angle with the antenna being so close to the ground, but the fact that I was able to work into the Middle East, India and Asiatic Russia all while barefoot, says something.

In the photo, it may look like the antenna is mounted to the chimney, which it is not. The mount is a modified non-penetrating roof mount for DirecTV-type satellite dishes with 8 low-profile concrete blocks acting as ballast. The lower mast is guyed to the plate that holds the blocks. The mast is then guyed to the 4 corners of the house and attached just below the rotor.

Oct 26

Shack Power Plant

Within the last couple weeks, my 50A Sorensen instrument power supply has been giving up the ghost. If I would draw anything more than a 19A load, the radio would start to output hash all over the band, generating a recklessly wide signal on CW. I must say, not very good amateur practice. Since then, I’ve been looking around for a solid power supply, even finding a nice 35A rack mount Duracom at a local hamfest for $75 which I ended up kicking myself days afterwards for not buying.

My search ended while browsing some eBay auctions and was instantly introduced to the Chinese OEM-type switching supplies used in various devices and appliances. The prices were very reasonable, 30A supplies for approximately $50. How good could they really be?

I remembered hearing something early on about switching power supplies causing interference in the shack, or generating hash so loud, it would cover-up the reception of weak DX stations. I immediately started my search on eHam to see if anyone else had attempted using similar supplies and expected to hear the usual horrific war stories. I was surprised to hear something positive on the subject with many hams having gone the same direction.

Here’s a link to one of such discussion threads:
http://www.eham.net/ehamforum/smf/index.php?topic=71829.15

With that in mind I found an eBay auction for 2 units and purchased both.

Upon wiring up one of the supplies with Anderson power-poles leads for output DC and the stripped back end of an extra computer power cord for the input AC, I was ready to unleash my abuse. Below you will see my findings tested with a power meter used for troubleshooting electric RC airplanes. The meter displays voltage, current draw, watts and amp hours and has been a very sought after tool in my shack. I highly recommend picking one up.

The first photo shows current draw on one supply leg of a 200w Kenwood TS-480HX at 100% duty-cycle, full output. Current draw measured was 19.5A.
The second photo captures nominal output voltage of 13.24V at minimal load. 1.2A was measured with the radio on and idle. This was the maximum voltage I could get out of this supply.
The third photo indicates the voltage drop with a 19.5A load. Output only dropped .25V.

OK so it looks like the supply will do 20A fairly easy. But will it output it’s rated 29A? I was able to draw the radio’s maximum current draw from just one supply. This is probably a bad idea but it does seem to do it.

In addition to testing the supply on the bench with an oscilloscope (which output a clean DC trace), I also tested it with the 480HX at full output on CW. Not a hint of RF hash was detected.

I’m only a couple days on the new supplies and I’m very happy with my decision to go this route. The RF hash issues with these types of supplies seems to be a thing of the past. Long gone as the technology has gotten better. However I did notice a voltage output difference between the two units I bought. One of them outputs 14.8V max and the other 13.6V. This doesn’t seem to be all that critical for this application, but could indicate a QC issue during manufacturing. I don’t know who made mine, since I can’t read Chinese, but similar supplies can be found on-line under the name Mega-Watt or other makes with the similar red lettering in the upper left-hand corner of the printed label.

One of the other things to keep in mind is that the terminals are exposed, a concern especially for the AC input terminals. I plan to mount mine in a 2U rack enclosure with Power-pole panel mounts, an AC switch and volt meter. That should address the electrocution any nosy cats who may exploring the shack.

Until next time, 73!
de N3QO
dit dit