A new QRP setup

Figure 1: My recent purchase (from QRP-Labs.com) is shown along with homebrew TTL/RS232 (click to enlarge).

I was looking at the kitsandparts “one watter” transceiver, but it’s a “one bander” rig.  It looks nice, but I decided that I wanted more flexibility in my new QRP rig.  So, I went to the QRP-Labs.com site, and found a nice way to get the flexibility I wanted.  In figure 1 can be seen the Progrock kit.  It went together in nothing flat and cost me $18 bux.  I think I’m a happy camper with this setup.  It provides a 25 milliwatt square wave signal, that I can run through an LPF to produce a nice sine wave, subsequently adding a series of amps and filters to achieve whatever power level I desire.  I’m thinking about using the 250 milliwatt driver amp that utilizes the BS-170, and follow that with a five or ten watt (selectable) IRF-530 homebrew amp.  The plan is slowly falling into place.  I purchased the Progrock because it’s cheaper than the complimentary VFO, and I’m going to control the frequency with my little homemade (pi2 powered) tablet anyway, via the serial interface (see below):

Figure 2: A terminal program is accepting the serial output from the progrock kit (click to enlarge).

I used a USB-to-RS232 serial adapter, which I connected to the homemade tablet, and then I connected the “RS232 end” of that to a homebrew RS232/TTL adapter, which feeds the QRP-Labs Progrock kit’s serial (TTL) interface. It works well, IMO. In figure 2 can be seen the terminal program, and the output from the little Progrock. Currently, it’s showing all zeroes in the frequency registers (all outputs are turned off).  Notice the staggered output.  That’s not the fault of the Progrock, but the result of the echo option I used on the Picocom command line.  It looks goofy, but I wanted to see exactly what I was putting out (and that’s it).

For the TTL converter I used the 2N3906/2N3904 PNP/NPN complimentary pair.  They are not rated for high output, but worked in my case.  Normally, only low currents should be involved, but a homebrewer should check to be sure of course.  I’m not specifically recommending them, but I had a hundred of each in my junk drawer  (they cost quite a bit less than the stamp to send them).  I used a 33 uF, 100V overkill capacitor (again, it’s what I had in my junk drawer).  The cap is not for power supply filtering, per se.  It’s included so that the RS-232 TX line’s negative voltage can be used to enable the RS-232 RX without adding another power supply voltage.

Figure 3: The whole ensemble: Progrock kit, TTL/RS232 adapter, Soft66/Lite RX, and homebrew tablet.

In figure 3, I’ve assembled the components, less the driver and power amps.  The Soft66 Lite SDR RX is described on other pages of this blog.  It works well with the SDR software running on the tablet.  It’s not the most capable unit in the world, but it matches the profile of the little Pi powered tablet very well.

Some notes are in order. First, when I assembled the Progrock, I took the advice of the kit makers, and put the 5VDC regulator on the progrock PCB instead of the synthesizer board. Then, I used the noise filter to derive the 3.5 volts for the synthesizer. After I put the two boards together, and BEFORE I mated the synth kit board to the progrock PCB header, I tested the voltages from the (a) 5 V regulator, and the (b) the noise filter. The filter voltage was 3.8 volts! A look at the Si5351A spec revealed that the maximum voltage is 3.6 volts. So, what apparently happened is that the carbon resistor tolerances all aggregated in the same direction, putting the voltage quite a bit over the 3.5 VDC mentioned in the kit assembly guide. So, I used a 68k “trimmer” resistor to get the voltage back down to 3.5 volts. I tacked it across the Progrock kit resistor R2.

Here’s another note of interest for Linux users. The QRP-Labs Progrock serial interface sends and receives only the carriage return code (#13) at the end of lines. The newline (#10) character that Linux users use in its place turned out to be a troublesome item to modify.  I could not figure out how to get the Minicom serial program to drop the LF and replace it with a single CR code. Finally, after futzing with Minicom for an hour, I installed Picocom instead, and that is what is shown in Figure 2.  It was necessary to use special command line options to set the newline ending character (the –imap, –omap, and –emap options apply for this task).  After getting the newline character corrected, all was good. Clicking on the graphic (above) will show it in large size.

Another small issue appeared when I constructed the RS232/TTL converter. The inexpensive USB serial cable I had in my possession turned out to have “not much ooomph” on the positive voltage side (“low” in RS232 parlance), and was insufficient to drive the NPN transistor into saturation to provide TTL output. On the positive swing, the voltage was not as high as the negative RS232 voltage was low (-7 volts, good enough in my case). RS-232 is supposed to be +/- 12VDC in order to be “compliant” IIRC, but none of the new stuff actually provides that much voltage anymore. I had to adjust the bias resistor on the HB adapter in order to compensate for the low voltage (although that probably means it won’t work when I use a more compliant adapter, LOL) I need a toggle switch for “adapter compliance level”.

OK, so enough of the setup stuff. How does it work? Very well, IMO! I hooked my CW keyer up to the Progrock, and connected the output to a dummy load (doubt if that was necessary). The progrock provides a pretty hefty signal, but I’ve seen a couple different power numbers in relation to the output.  Most recently, I think I saw that it was around 25 milliwatts.  That’s much higher than other typical devices in this genre of functionality.  In any case I used the keying scheme described on the QRP-Labs discussion site, wherein the default bank of Clk0, Clk1, Clk2 are all zeroed (turned off), and a “bank 1” Clk output was “keyed” via the bank-selector pins. I noticed no chirping or clicking. It actually sounded pretty good – maybe nice enough to hook up the antenna and bang out a 25 milliwatt CQ.  But nah – I decided to wait until I (at least) have the driver built!

OK, so I was operating at 7 MHz. The signal was still good and clean at 10, 14, and 28. But – at 50.1 MHz, it was drifting somewhat. I should mention that I don’t have the “crystal oven” version of the progrock or VFO. For six meter work, it’s gonna be needed! The crystal oven adds only about $7 or $8 bux, and is HIGHLY recommended if one is going to exceed the upper limits of the HF spectrum.

Read More …

Note: the soft66 Lite hardware is a product that is sold on a Japanese website (http://zao.jp/radio) – and is not affiliated with this site or author in any way.

Note: Kitsandparts at kitsandparts.com is not affiliated with this site or author in any way.

Note: QRP-Labs at https://www.qrp-labs.com is not affiliated with this site or author in any way.

Note: The Raspberry Pi Foundation makes the Pi2 referenced in this post. They are not affiliated with this site or author in any way.