Sunday, November 18, 2018

2018 ~ The Year of SSB Transceivers

A New Transceiver ~ Dual Filter Rig

11/23/2018 ~ A case for "Troubleshooting".

Since much of the prep for Turkey Day are now over thus I had a few spare minutes to work on some radio projects. I needed to use a bigger wattage soldering iron to "weld" some copper PCB cases together. When I plugged in the iron purchased from MPJ in FL, the ON LED  was intermittent and the iron would not heat. This required a disassembly of the controller since I knew the iron itself was good. 

What a bear to get that case apart but I did manage to get the unit out of its case. Actually the soldering control box is kind of cool as it has a detachable 3 prong cord and built into the unit receptacle is a fuse assembly and upon checking the fuse discovered there was even a compartment for a spare fuse. The fuse checked good so that was not the issue,

I "jiggled" the wiring from the transformer to the board and saw the LED flash on/off. Boom, a loose connection. But interesting was that the PC board was burnt around where the wire from the transformer to the board and the solder connection looked like it was a cold joint but much evidence of burning. The wire connection evidently was loose and arcing which would cause the burning; but it would have to get awful warm to cause the solder to have that cold joint look. All other wiring to the board which went to the actual soldering iron itself looked "kosher". 

Thusly I cleaned up the solder joint and re-soldered the wire -- and before installing  back in the case tested the power ON LED -- we were good. So everything buttoned back up and I installed the soldering iron. With the fine tip point -- it worked AOK. But when I installed the 3/16"  wide screwdriver tip it would not get hot enough to melt thicker solder.

So there must be more to the reason behind the loose/burnt wire and now back to opening up the case again to do more trouble shooting.

You cannot have too many soldering irons and I have five of them. Two look like brothers --one was purchased from Marlin P Jones (detachable cord) and the other from Circuit Specialists. Same case but different logo; but the one from Circuit Specialists has an integral cord and no external fuse.

So I am at a loss --both units were under $30 and the one I use for hardware connections is an Xtronic which cost me about 2X that -- it is nothing to write home about. 

Anyone have any soldering iron recommendations?

73's
Pete N6QW




Only 6 weeks left in 2018 --need to get cracking if I want to build at least three more rigs this year. Keep in mind that I reuse many of what I call standard building blocks. Thus using these standard circuits enables fitting together what I call electronic Lego Blocks. Having the Digital VFO/BFO vastly reduces the prototype time.
A Bit of WSPR & FT8 on a Totally Homebrew Station. 11/20/2018



Earlier this year I cobbled together a dual filter rig and sort of just put it on the shelf after getting some early results. By dual filter this topology means that there are separate LSB and USB Filters and to switch sidebands the BFO frequency is fixed and the filters themselves are switched to either sideband. Drake used this method as well as many commercial manufacturers. This has some charm in that you only need to contend with coming up with one BFO frequency.

Late in 2017 I spotted a surplus crystal filter board that contained a set of the filters as well as a 3.6 KHz AM filter and the center frequency is 9 MHz. The surplus board is sold on eBay by a firm in Israel. The filters are made by KVG (Germany) and are quite excellent. The specifications for the filters appear as a later photo. See the link below --now about $38 delivered. I paid $34

https://www.ebay.com/itm/KVN-Kristall-Crystal-Filter-BP-9MHz-XF-9-SSB-AM-Receiver-RF-HAM-Radio-Amateur/141008771238?hash=item20d4c718a6:g:sNQAAOSwn-tZEv8H:rk:36:pf:0

The amazing thing is that I got the board shipped to my QTH for $34. The filters are installed on a circuit board that contains many useable parts --some are unobtanium! These boards I think are still floating around and the price is roughly the same. I decided to make the rig a multiband rig but currently only the 40M BPF and LPF installed. I am impressed at how good it sounds.

The line up is as follows:
  • It is a bilateral rig that uses the Plessey amps (2N3904/2N3906) ahead and following the dual filters. For the Rx Tx Mixer I used an ADE-1 and for the Product Detector/Balanced Modulator, a second ADE-1. To switch sidebands.I used diode steering. It works very well. 
  • The audio stage is a 2N3904 and LM386 and the Microphone amp is a 2N3904
  • The Rx RF Amp and the Tx Pre-Driver use the same stage which is a pair of J310's configures as a Dual Gate MOSFET. A couple of relays steer the signals through this single pass amplifier. 
  • The Driver stage is a single 2N2219 and the final is an IRF510. 
  • Recently I built a an Arduino/Si5351 into a repurposed copper PCB box that formerly housed a RF linear amp. This now is the control panel.
  • The rig puts out 5 watts but I can put two RF linear amps in line for a whopping 600 Watts +. With that configuration I worked EA3JE, Lou in Barcelona in the middle of the day on 40 Meters. Lou said I was 10/9. So this rig will run with the big dogs. Then today I worked Hawaii during the SS contest running QRP. My antenna is a sub-optimum 40 Meter dipole on a postage stamp sized lot.
  • The rig uses two 4X6 copper PC Boards that are stacked on top of each other. The  bottom board has the filters, the Rx Tx Mixer and PD/BM. The top board has the Audio/Mic Amp, the single pass J310's, the BPF , Driver stage and the TR relays
  • External to the stacked boards are the Control Panel, the IRF510 and Heatsink and the LPF.
  • Controls Include, MOX, Tune Button. USB/LSB Select, Band Select, Main Tuning Audio Gain, Microphone Jack and Audio Jack.
  • This rig like others I have built provides a 988 Hz Timed Tone for Tune Up. Sure beats screaming Hola into the microphone. A screen display of TUNE in Red is shown when in the Tune Condition. How cool is that?
  • This is not a compact rig but does facilitate experimentation which is always a desirable goal.
There is just something amazing about fielding a totally homebrew station and this new dual filter rig is just that -- totally amazing. But the icing on the cake is that you can run the digital modes --that too is also homebrew. Many in our hobby can do also do this same feat. So my skill set is not totally unique or one of a kind but it does speak volumes to learn the "how". It is simply too easy to whip out the plastic and you are there. There is also a tribal knowledge factor --I have been doing this for 60 years which flies in the face of some response such as I am too old. You can do the math -- I got licensed when I was 17. Back in the old days you had to build much of your own station and plastic was not in widespread use. So make yourself a New Year's resolution to totally homebrew a rig for 2019. Sideband would be nice but CW will suffice.










73's
Pete N6QW

PS Still on the bench are at least three more transceivers projects including one not unlike what Peter Parker, VK3YE built as a single channel 40 M SSB rig only on 7.2 MHz. I am noodling a really small, trail friendly, portable rig using surface mount technology and no VFO. Thanks Peter for the inspiration. Another is to rip out the crystal switched VXO from a  the 17 Meter rig and install a Pro-Mini and 8X2 LCD. The 3rd is a mystery project. Can this all be done in 6 weeks?


2 comments:

  1. Hey Pete, Along with some fellow hams, I do volunteer work at local schools, at the middle school and high school level, where we teach students about electronics. In our high school group we have a few students interested in receiver construction. I've produced PCBs of individual RF blocks from your Simpleceiver design. The students construct the blocks. We test them, and then connect them in various receiver configurations. The students are learning a lot about RF blocks and receivers.

    I'm currently working on a configuration which will use three PCBs (RF Stage, IF Stage, and everything else) designed to mount in an enclosure with an Arduino and LCD so students can complete a nice superhet receiver they can keep and take home.

    Today, as I was using LTSpice to look at the input impedance of the RF amp under different load conditions, I noticed something interesting. Even though your designs show a 2.2K resistor between the FET gate and ground, and your impedance matching is for 2.2K, the simulation indicated the input impedance to be half that, with a 3dB increase in amplifier gain without the 2.2K resistor. I took one of the RF amps to my workbench and checked the gain with and without the resistor. The testing confirmed a significant gain increase without the resistor.

    From what I can determine, the input impedance is 2.2K with no resistor. Adding the resistor drops the impedance to 1.1K, thus causing a mismatch.

    Of course, it's entirely possible that I'm overlooking something. If so, could you set me straight?

    Thanks for all your radio design postings!

    ReplyDelete
  2. Hi Will,

    Thanks for sharing what you are doing with the Students. Chunking things with various blocks enables a better understanding of what is happening inside the blocks and also facilitates the ability to make changes.

    Relative to your comment about the 2.2K. I used that value to provide a purely resistive load to the matching network. Now if you remove the 2.2K, two things happen in the simulation I am using. Firstly, the peak of the amplification curve drops about 2 MHz lower in frequency and 2 you do see a gain increase. So if you were to use this on 40 Meters you would need other circuit adjustments to boost the frequency at 40 Meters not 60 Meters. The resistor was added also for the DC biasing to Gate #1. Now if you made the resistor 10X which would make it insignificant as a part of a parallel circuit you would get the gain boost. I see it going from about 9 dB with the 2.2K to about 16 dB with a 22K. But you still have the issue of a peak at 5 MHz versus 7 MHz.


    LT Spice is a wonderful tool and lets you see many things -- If you adjust the other parameters so you get peak gain of 16 dB at 7 MHz then have at it.

    I do think I saw the 2.2K in other designs (W7ZOI) so it may not be original with me. But congratulations on using the tools and asking the questions -- that is what the learning journey is all about.

    73's
    Pete N6QW

    ReplyDelete

2018 ~ The Year of SSB Transceivers

What it takes to build a SSB Transceiver? I could flippantly answer the question by saying a lot of luck, lots of parts (and/or lots of...