Thursday, April 6, 2017

Hot Water BITX 40 Hack

Fred's idea really resonated with me.   My first SSB rig was an HW-32A, the 20 meter version of the rig shown above.  If -- as I suspect -- these rigs are anything like the HW-101, they are not aging well. Heath's drive for economy resulted in rigs that don't hold up to well over time. I remember the sound of the  plastic HW-101 dial clutch cracking when I pushed the button.

BITX40 Modules to the rescue! Put a mono-band board inside an old mono-band rig.   There are a lot of possibility here.  Some ideas:

-- Put that Heath VFO to use.  Maybe convert it to solid state.  Or just put the LCD from an Si5351 in the window (Pete did this with an HW-101).

-- Get the S-Meter wiggling.  

-- Keep the final amplifier circuitry in there and let the BITX drive it.  This will give you a QRO option.  Bill N2CQR 

Hello Fellows,
Attached is a picture of my BITX-40 V3 adapted to a Heath kit Single Bander HW22. This is a work in progress but what a neat way to bring an old boat anchor into the present.
 
The only parts of the HW 22 used were the front panel and case and knobs. Modifications yet to be  incorporated include: AGC , a USB port on the front panel to access the Arduino, and a PTT/CW mode switch.
 
I enjoy your pod cast and web site…Best of 73 KC5RT.

Wednesday, March 29, 2017

ND6T's Forward and Reverse Power Meter for the BITX


RF Monitor
by Don Cantrell, ND6T

In response to several requests for a VSWR and output power monitor I have developed a simple circuit to be easily added to the BITX (or any transceiver) and some software to get it working. I used a 20 dB directional coupler design, sometimes called a “Stockton bridge”, that is simple, broadband, and requires no adjustment.

The sensor circuitry is mounted on a small piece of printed circuit board that is held in place under the mounting nut of the BNC antenna connector, replacing the ground lug that was originally supplied. The (brown) wire from the 2-conductor “ANT1” plug feeds through the center of one toroid core (T1) and solders to the BNC jack, the other (black) solders to the new board's ground plane. Two new wires connect to the Raduino plug for reading and display of the results.

I began with a square of un-etched printed circuit board stock, one inch on each side, and drilled a 3/8” hole near one corner to fit over the antenna jack. Leaving enough room for the mounting nut, I glued 4 small pieces of PC board as solder pads. The two transformers are T37-43 ferrite cores wrapped with 10 turns of AWG#24 wire in a single layer. The 50 ohm terminations are ½ watt 51 ohm resistors. I hand-selected two of the ones closest to 50 ohms but this is not critical (what's an ohm between friends?) ¼ watt will work fine at these levels. I used two 1N34A diodes for detectors.

I mounted all of the components upright to save space. A short jumper wire feeds from T1 secondary through the center of T2 to the forward detector as that primary winding. Note that you do not wrap this around the core, just pass it through like the antenna lead on T1.

Don't have any #24 wire? Use whatever you have that will fit 10 turns. Don't have a T37 core? A T50 or even T60 will work. It is best, however, to stick to 43 mix ferrite unless you adjust the turns accordingly. Capacitor voltage or temperature coefficient is not important. Got lots of room around the antenna connection on your build? Then go large with the PC board and give yourself some space for component mounting and ease of connection.

Testing is simple. If windings have the proper polarity then, when you transmit into a dummy load, there should be a couple of volts DC on the “Forward” solder point and a very low voltage on the “Reflected” pad. The coupler works the same in both directions, meaning that if it were built outside of the BITX, then you could reverse it and get the same readings by switching voltage pads. A nice check of operations. 7 watts on my build resulted in a bit over 2 volts volts on the output, well under the maximum 5 volts allowed on an Arduino® A/D input.

I included some formulas in the software to average the peaks during SSB operation and displaying it every 3 seconds. The results are also displayed after a half-second key down in straight-key mode. The power reads out as watts in both forward and reverse directions. As usual, the Arduino sketches are available free by emailing ND6T@arrl.net.

Not using a digital readout? Just attach meters and adjustment pots to the output connections. If you want just one single-movement meter then use a switch and a trim pot for each switch position.

de ND6T

Tuesday, March 28, 2017

ND6T BITX CW Update


BITX CW Update
By Don Cantrell, ND6T

Since the last post (February 25, 2017 “Putting the BITX Raduino on CW”) I have made a few changes in the sketches, added a variable keying speed control, and added a Straight Key mode. In the ensuing weeks it has performed nicely on the air, superb reports from everyone.

The primary reason for the initial sluggishness in calling up the original keying routine was that the display updating necessitated a sizable delay to keep it from flickering. By moving the display update out of the loop() and putting the CW into a “while()” structure made the delay essentially disappear. It also cleaned up the code, modularizing it to make it easier to re-configure and to steal whole sections to use in other sketches. I decided to use iteration counts (how many times the subroutine ran) to create the time-out to return from CW transmit to receive. This is called “semi-break in” operation and is used on many rigs. This is still primarily a 'phone rig, after all. I was reminded about how fast the Arduino works by the number of iterations it performed in 1.5 seconds...100,000! Zoom! Yes, it read the paddle inputs 100,000 times in one and a half seconds.

The next big change was a speed control. I am not a fan of menus. I want to turn a knob and go. So I drilled another hole in the front panel of the BITX and mounted another 10K linear taper pot. I wired it like the tuning control: +5 volts to one side, chassis ground to the other, and the wiper to an unused A/D pin on the Raduino. Since the new speed control was right near the tuning control, I just wired from that for the end connectors.

The CW function reads the value and converts directly to Words Per Minute. The display shows the two-digit speed at the end of line 1 right after the “Idle” indicator on my program. I have it set for 10 to 35 WPM, pretty much my normal range. It can be changed to cover whatever range the operator would like by a couple of simple changes in the code. That code is still extremely short, small, and in simple Arduino.

I am not a person that shuns Straight Key mode, though. In fact, I rather prefer it. It is simple, understandable, and has character. Much like a persons' hand-written messages. An operator's fist is as distinctive as his call. And who can send CW slower than 10 WPM on a paddle?

With the current sketch, turning the keyer speed control to less than 10 WPM replaces the speed indication with “SK” and directs the program flow to the straight key function. Again, after 1.5 seconds of no key closure, the BITX returns to receive.

The only glitch occurs if the straight key uses a 2-conductor plug or has the sleeve and ring conductors shorted and is plugged into the 3-conductor key jack. Do that and the Raduino sees it as a request for a “dash” from paddle. This is handled by reading the speed control. If the control is set for “SK” then all is well. Leave it on a keyer speed position with such a straight key plugged in then you can expect the BITX to start transmitting “ DAH DAH DAH...” until you realize your error.

Although I was initially worried about key clicks due to the fast rise time of the keying envelope I don't see much difference from my other CW rigs. My Icom 746pro is not bothered just 100 Hz away from the 50dB over S9 signal when using a 50Hz receive filter. No one I have worked notices a problem. One operator using my conversion thought it had some chirp but it turned out to be receiver overload instead. Not the BITX, another receiver, relax.

Dmitry, EI3JQ, reported that he had to use a lower value series resistor at the signal injection point at C132. He says that the 470 ohm resistor resulted in low output. Further reports are still pending but I expect that different builds will show variation. It is my opinion that the CW output should be no more than the SSB PEP level to maintain signal purity and lessen component strain. The average power difference is nearly 10 dB and should be taken into account. Dmitry had much lower power, though. He reports nice CW contacts both in Europe and Russia now.

I also re-wrote the sketches to reduce the number of synthesizer and display refreshes. The newer versions of the etherkit Si5351 libraries ( v 2.0.2  and v 2.0.3) shut off the synthesizer outputs for a brief time when programmed for frequency. This creates a clicking in the audio. My earlier sketches were constantly updating so it was particularly irritating. Jerry, N5KYD, found this problem. I suggest using v 2.0.1 until the library is further corrected but my newer sketches have profitted from the experience.

If anyone wants a copy of the current sketch with CW, for either the 40m Raduino BITX or the 60 meter version, just write me at: ND6T@ARRL.NET  for a copy. There are changes made almost daily, lots to improve. The fun is in the innovation, after all!

de ND6T

Thursday, March 9, 2017

A Band Scope for the BITX




The use of an RTL-SDR dongle with a discarded Android phone as the display seems to me to be a very BITX-appropriate use of road-kill technology.  Bill N2CQR

Thoughts and experiments on BITX Band Scope
by Ken Marshall G4IIB

I have explored several very simple ways of implementing a band scope on my BITX. To test the concept I am using HDSDR on a PC and an HF converted RTL SDR dongle (a la Sprat 162). We can turn the sound off on our SDR software as we are only interested in the display. I have also tested it in demo mode on an android phone using SDRTouch as seen above.

Method 1 implementing a band scope on the first IF stage (output of Q2). This gives a classic tunable Pan-adapter. You tune your SDR software (HDSDR or SDRTouch) to the IF frequency 12Mzh and you will see a band scope and waterfall display.  You tune stations in the normal way on the BITX and you will see that once you resole the LSB signal on the BITX it resides within the 12Mhz pass band near 12Mhz on the waterfall.  See below you will also see that the lower portion of the 40M band is to the right note the CW section 1260 – 1280 and the high end of 40M is to the left  this is inversion due to my use of high side vfo. This will not be he case if you are using the conventional low side VFO ( the waterfall will change if you use high side VFO to change to USB as implemented on the Raduino using the fixed BFO crystal, if that’s what you use). What you see on the scope is not what you get in that every thing is referenced to 12Mhz you have to do a conversion in your head to know where you are within the 40M band, but it does show you the activity on the band which you can tune to via the BITX.

The next image shows the mess the waterfall gets in to when you tune the band. The trick when tuning is to ignore the waterfall and concentrate on the peak in the band scope of the signal you want to listen to as it moves into the 12Mhz pass band when tuning the BITX. 




As previously indicated the hardware mod itself is very simple, tack a 10 or 20pF cap (whatever low value you have) onto the crystal side of C23 and connect this to your dongles antenna connection via screened lead.

Method 2  implementing a band scope on the RF amp stage (output of Q1). This gives an untunable Pan-adapter/band scope. You tune your SDR software (HDSDR or SDRTouch) to 7.1Mzh and you will see a band scope and waterfall display of the whole of 40M.  You tune stations in the normal way on the BITX the band scope does not move as you tune all of the signals you see are in the correct place on the band. If you see a signal on the band scope at say 7140 tune the BITX to 7140 to listen to it, the band scope does not change. As the main purpose of the band scope is just to give a visual indication of what activity is on the band you may find this approach simpler to use.

Again the hardware mod itself is very simple, this time  tack a higher value cap 10nF onto C13 and connect this to your dongles antenna connection via screened lead.

The main downside to methods 1 & 2 are the amount of spurs that can be seen in either of the displays as above. This is a known issue with these cheap RTL SDR dongles and not easy to overcome. However, most of the spurs seem to be generated within the BITX. This is hardly surprising given that I'm running an arduino, which has its own internal clock, which is driving an SI5351 producing VFO and BFO frequencies, harmonics from these and those produced by the RTL devise itself will be generating and mixing all sorts of other harmonics. Contrast the above with the SDR on its own antenna not connected to the BITX. Hardly any spurs see below. This would be another option (method 3) to use the dongle with either a PC or android device as a band scope using a separate RX antenna.

 I have loaded SDRTouch on my android phone and it works but only for 60 seconds in demo mode. Not really long enough to test and fully evaluate the functionality. I have to pay£7 for a key to use the full spectrum functionality. I'm not sure at this stage which android devise I want to run it on or the use I'll make of it. But I may consider it in the future. This is why most of my testing has been carried out via HDSDR on a PC. I have tried other android SDR apps but SDRTouch seems to be the only one to offer I or Q channel inputs necessary to operate HF modded dongles.

Whilst methods 1 & 2 work they would benefit from further experimentation. Reducing or eliminating the spurs would be nice but difficult. I have since found that screening the Arduino and using coax instead of just screened lead to connect to the RTL helps. As it stands even with the spurs it still gives useful band information. A band scope found on $2000 plus rigs it is not! But for the outlay of $13 for an SDR dongle and an android device it could be a useful addition to the BITX.

In summary, both methods are very easy to implement but I feel that method 2, although not a classic tunable band scope, is probable the easiest option from a user perspective. Try it on your PC or android and see what you think.

Post Script check out these V3 dongles from the RTL-SDR.com people. They have improved screening, spur suppression, lower noise floor and are already modded for HF. It's only 19 of your American bucks.
http://www.rtl-sdr.com/new-rtl-sdr-blog-units-now-available-in-store-hf-via-direct-sampling-software-switchable-bias-tee-less-noisespurs/

Saturday, February 25, 2017

Putting the BITX Raduino on CW


Clean CW on the Raduino
 by Don Cantrell, ND6T

Although the original Raduino v 1.01 sketch included a CW mode I wasn't pleased with the signal produced from the side tone insertion into audio. Even when filtered to where it sounded acceptable there were multiple artifacts accompanying the tone. Since I like a 700 Hz tone there were three additional harmonics spread across the spectrum within the 2.8 KHz spectrum. These are attenuated by the degree of filtering but unless I used active filtering they were quite evident. Additionally, there was the suppressed carrier and the opposite side band both showing at several milliwatts. All this would not be detectable if my signal was quite weak, but there are many times when nearby stations will be annoyed by these “companions”.


I decided to use the CLK1 output of the Raduino module (P3 pin 11) inserted, at operating frequency, into the RF Power Amplifier predriver (Q13). I used a 470 ohm resistor soldered to C132 (either side, CLK1 has a blocking capacitor) fed with a shielded RG-174/u cable. I grounded the shield at the synthesizer at P3 pin 12 but left the other end open, dressed back ½” , and covered with tubing. The 470 ohm resistor throttled the CW output to 5 watts. If you want to chance it, a 220 ohm resistor at that point will output 10 watts. I prefer 5.

 The T/R relays are keyed by placing a 2N7000 NMOS transistor from ground to the PTT pin on the microphone jack. Source to ground, Drain to PTT, and Gate to P3 pin 1 of the Raduino.

Another 2N7000 is placed across R44, the base of Q4 to ground. Source to ground, Drain to the base side of R44, and the Gate tied directly to the gate on the 2N7000 T/R driver that was just installed. This transistor disables the post-mixer transmit buffer in order to eliminate the remaining carrier and anything coming in from the microphone.

The result is a clean and clear CW tone.

In my software I use a side tone that matches my transmit offset. A 700 Hz offset will produce a 700 Hz side tone, making it easy to spot my transmit signal in the receiver. This side tone appears on P3 pin 2 on my build and is injected into the speaker jack through a 1 uF capacitor. Those users who have chosen to remove C113 from the audio output amplifier can use that capacitor: Just lay it on top of the jack and solder one end to the “tip” connector pin (the brown wire). Want a different level? Then either change the capacitor to a different value or put a resistor in series with the 1 uF. Another variable resistor on the front panel, perhaps?

I wrote a primitive little iambic keyer routine and placed it in a “while” loop. The structure is timed by a simple iterative count and jumps back out to normal operation after about one and a half seconds of inactivity. That is the “semi-break in” time out and is changeable in software to meet your needs. The key inputs are fed into P3; pin 3 for the dits, pin 4 for the dahs.

Two transistors, one resistor, and a capacitor. Less than $1 USD in parts!

The CW routine is called by holding the paddle to either side. The relays are operated and the fun begins. Speed is set in the start of the program as a variable, “wpm”, and is fixed on my simple sketch. You can set up a simple pot (just like the tuning control) to generate a number to be directly scaled to the word-per-minute rate and place that control right on the front panel for speed control. Place the WPM rate on the second line of your display. Class!

You could do the same thing with the offset/side tone frequency. Go crazy and amaze your friends!

I can provide a current copy of the 60m sketch that includes the CW program. It is written in simple Arduino, no fancy C++ calls. I like simple. Also included is the added tuning of 5 to 6 MHz. Nice to calibrate with WWV, listen to trans-oceanic aircraft and short-wave broadcasts. Write me at ND6T@ARRL.net.  An adaptation for 40 meters will also be available when I get to it.

It certainly serves as a successful proof of concept. It is not, however, a great CW transceiver. Yet.

*       Activation of the CW function is awkward and slow. It is delayed by the iteration of  the main loop, including the display. Solvable by an interrupt routine, perhaps?

*          It needs a straight key mode.

*          Keying waveform is too abrupt. Perhaps an keyed integrator in the PA power plug circuit?

*          Receive bandpass it pretty wide for CW. Audio filter?

Plenty of room for improvement and more modifications. Isn't that the whole point of the project? Incremental improvements and improvisations, learning (both hardware and software), collaboration and open-source hardware and software. These are grand times.

de ND6T

Tuesday, February 21, 2017

Cap Stack Hack: Putting the BITX40 on 60 Meters


Putting the BITX40 Module on 60 Meters
by  Don Cantrell, ND6T

I have been enjoying the BITX40 with  the AD9850 DDS VFO but when the Raduino was announced I was overwhelmed with curiosity. I ordered a new BITX complete with Raduino. What would I do with another 40 meter BITX? How about 60 meters?

60 meters in the U.S. is pretty limited, currently only 5 fixed channels and limited to 100 watts effective radiated power, and so it is seldom addressed in commercially built gear. It has good propagation for nearby communication and there is a possibility that a new segment (not channels!) will soon be authorized for QRP operation. The BITX would be perfect since it is easy to program for new allocations.

While waiting for the new BITX I considered conversion strategies by using the AADE filter modeling program. First, I modeled the existing 40 meter RF bandpass filter and then modified it for 60 meters. My first idea was to change just the series capacitors in the original. It worked!

The nice curve to the right (in black) is the original RF band pass response. Changing the three series-tuned capacitances produced the response to the left (in red). Yes, it is twice as broad, but that is what I wanted since it would include WWV and a number of frequencies that I enjoy.

The conversion does not require removing any hardware. Just add four common value capacitors! Parallel the three 100 pF capacitors (C2, C4, and C6) in the RF bandpass filter with 100 pF capacitors. I  just stacked them. The fourth additional capacitor is a 220 pF capacitor across L7 (in the transmit low-pass network) to attenuate to second harmonic (now > 50 dBc). Voltage ratings are not a problem.

Stacking is done simply if you have some 1206 sized capacitors. Just sweeten the solder fillets on each end of the target capacitors with just a touch of extra solder. Place the additional part atop the target part. Hold it down while re-heating the solder at each end and you will see the capillary action pull the solder up onto the ends to complete the joint. Ta-DA!

No SMD capacitors in your junk box? Just use ceramic disks, then. The 220 pF needs to have leads  in order for it to reach across the L7 terminals and it goes underneath the board. You can use a ceramic disk like I did, but a silver mica or any other RF type would work well.

That was too easy. The only thing left to do is to load new operating software and put it on frequency.

I wrote an entirely new sketch for the Raduino. Instead of tuning a band, we now have (currently) just five channels. The new program has you tuning the knob to the far clockwise end to start the channels  changing from #1 to #5 and then starting at #1 again. I will add other channels for WWV and the broadcasters to suit me and will institute tuning when the allocation is approved in this country. This way I don't need to modify more hardware. As it now stands, I can just turn the knob to where it slowly  scans all of the channels and walk away until I hear activity. To stop channel selection, turn the knob counter-clockwise anywhere out of the channel select zone.

I will provide the 60 meter sketch to anyone that is interested free (of course) if they will write me at ND6T@ARRL.net unless Bill or Pete can think of another way for me to post it. You can use it as a basis for any kind of experimental platform that you would like since it is very small and simple.

To put it right on frequency it would be best to use a frequency counter. This is due to the variations in BFO frequencies and the reference oscillator on the Raduino module. You could do it without  equipment by “change and re-try” but a counter (even a borrowed one) would be easier.

This calibration is the subject of a previous article.

de ND6T

Monday, February 20, 2017

VU2XE's CAD files for a BITX Box


Kiran VU2XE built a nice BITX box using newly developed CAD skills. 
He suggests that you could take these files to a local laser/CAD shop and have them reproduce  box in aluminum. 

Here (I hope!) are the CAD files. 

http://soldersmoke.com/GP_Metalbox_top.dxf

http://soldersmoke.com/lcd_Metalbox_bottom.dxf