Friday 3 November 2017


For the last  few weeks I have been busy twiddling  with my little QRP BITX rig. Of late, I have come across some problems and have devised some mods to these problems  and I feel these worth sharing  with  you.  

1. IMPROVED MIC AMPLIFIER: I felt that my BITX suffers from low microphone  sensitivity. You have to speak quite loud to get the signal properly modulated. A crass examination  of the microphone amplifier revealed that values biasing components seem to be little inappropriate and need to be recalculated.

Values of R126 (collector load of Q12), the mic amplifier  transistor and R 123 (emitter swamping resistor) are thus recalculated  and are replaced with a resistance of 4K7 for R126 (in place of its original  value of 1K) and R123 is replaced with  a value of 10 ohms ( in place of 100 ohms). You can spot these components near the left edge of PCB. 

The results of this simple modification were immidiately recognizable and were as expected. The rig now has required  microphone sensitivity and transmissions  are reportedly better.

2. CLICK FREE BITX MOD. : One annoying thing about BITX I felt is relay clicks and clicks caused by relay K2 in audio line.  Consequently  I decided to implement solid state circuitry to replace both relays used for TX/RX supply switching  and antenna changeover. The supply switching  circuit  is built on a pigmy  board as shown below :

During RX transistor conducts andRX stages get supply to work. During this condition Q3 grinds the pin 14 of K4 (Ex)  to ground the drive to pre-driver stages of the RF power amplifier. When PTT is pressed Q1 turns off and Q2 conducts to extend supply  to TX stages. During this period Q4 conducts and it's collector grounds the top (live) end of volume control causing the RX to mute. C1 and C2 ensure reliable switching  function even if a spurious RF radiation picked by the base of these transistors. It is necessary to remove capacitor C11 connected to the base of transistor Q1, for this MOD. as the antenna changeover function is performed by the following circuit :

I used an axial type moulded inductor readily available  in local market  for this MOD. However you can wind Your Own on a toroid available  to you.  The values of the components are not critical and are readily available. The results of this  MOD.  are really  satisfying. A professional  click free QSK experience.

3. AN R.F. BASED AGC MOD. :    BITX is a good  portable  rig  but its major handicap is that it lacks AGC. Though an AVC (automatic  volume control) circuit can be used to compensate it and is available online. I decided to include a more convincing RF AGC circuit to my BITX. Primitively it was thought to apply AGC through IF amplifier stages in a conventional design topology. So I thought  to put a transistor or a FET in the writer circuit of IF amplifiers as shown below.

The AGC voltage applied to the base of Q2 will cause to vary the current t in its collector circuit thus controlling the overall IF gain. But it was thought afterwards  that instead of course controlling the gain at IF level it would be better to control it before the mixer that would be an incentive considering the IMD performance of the mixer and all stages following it.  Consequently the following topology was initially thought to provide an effective AGC at RF stage. 

But in this case reducing the base voltage of RF amplifier can adversely affect the overall IMD performance of this stage. So after a detailed  analysis I reached the conclusion to use a FET to replace the Q1 based RF amplifier of the original design.  The resultant design is as under:

Audio signal is picked up from the hot end of the volume control and is amplified by transistor Q1. An AGC signal is then generated through  diodes D1 and D2, which is then used to control the gate bias of RF amplifier and thus its gain. The entire circuit is built on a small pigmy board and is installed close to the original BITX board. As the output of amplifier is about 440 ohms, you need to adjust primary turns to about the double of turns of other windings for a proper match. The results are as expected. Tuning through a crowded band full of 59+ signals is no more a bane. 

Wednesday 23 August 2017

Two Essential Add-ons For Bitx and Other QRP Rigs.

In this post I am going to discuss two very simple but very essential add-ons for Bitx and other similar type of QRP sideband rigs. These two add-on modules are the recent addition to my XENA and BITX rigs that have already proved their worth during portable QRP operations. The first one is of course a simple speech processor. It can give more punch to your transmissions and more QSO's are certainly assured for you.

A SIMPLE SPEECH PROCESSOR:  A simple speech processor can enhance your readability on the receiving end by as much as two S-points and is an indispensable addition to any QRP voice rig. There are usually two types of speech processing techniques. One using limiting through compression and other through clipping. The latter is usually known to give better results on both AF and RF. The present circuit uses this technique. The signal from the single stage mic amplifier is routed through this module. The first stage gives it necessary amplification and the signal is clipped using a pair of back to back diodes. I used germanium diodes as these were available in my collection but BAT 54 or similar are likely to give better results. The signal is then filtered using an op- amp based sallen key filter for harmonic suppression and is then amplified and is routed to the diode DBM. The circuit can be use with almost any SSB/DSB rig. For use with bitx you should use 47K resistance for R1 or use a potentiometer at the output to get proper level of audio for modulation.  The circuit is very simple and is given below:

I have chosen an LM358 for being cheap, readily available and for its smaller footprint than discrete devices, to aid smaller design. You can play with the values of C1 and C4. Try a value between 0.0033uf to 0.0068uf for C1 to suit your voice quality. Generally the least bass input to the DBM is the key to best readability. Try 1nf for C4 for much less splatter across the transmitted bandwidth.

The alignment is simple. An oscilloscope can be indispensable but if you have none don't worry. Measure the average AF signal amplitude at DBM input using a sensitive AF meter. Now switch to compress mode and set the viper of R8 all the way to ground. Now adjust R4 for an equal and almost similar amplitude reading at the input of your DBM. Go on the air and call a friend. Adjust R8 now for best results......and you are all done.

A COMPACT ANTENNA TUNER: A simple and compact antenna tuner is made of just two components; C2 and L3. L3 is a half watt resistance sized moulded inductor of about 12 uH. For 30 and 20 meters an inductor of about 6.8 uH would be sufficient and for twenty through fifteen meters just 4.7uH suffice. C2 is a common BC type variable capacitor and is not hard to find. I usually use a half wave throw away wire for my portable QRP use as it is very convenient to carry.

This compact tuner can tune both high and low impedances very effectively and is small enough for portable use. I have included N7VE's SWR bridge for ease of tuning. The entire module is small enough to be built and accommodated within the bitx cabinet.

Sunday 16 July 2017

VXOs-Simple Minimalist's Signal Sources.

With the recent advancement of technology, the availability and accessibility of stable signal sources has become within reach of an average radio home brewer. The newer versions of super stable and very dependable synthesizers based on both phase lock loop and direct digital synthesis are available for a moderate price. In addition to their accuracy, they usually come with an ornamental frequency display as well, to decorate the front panel of your rig.  

But for the minimalist QRP enthusiast X-tals are usually considered the cheapest stable alternatives. In practice though; they are undoubtedly cheap and stable but they do not offer the needed agility. Recently some good VXO designs have been published by some ham friends those offer the desired agility to work.  

In almost all of the projects I described, I use VXOs and I preferably advocate them as a low cost Scrooge's first choice. In this post I describe two ceramic resonator dual band VXOs, those can be adapted for X-tal operations as well.

1. BIPOLAR VXO: The bipolar VXO circuit uses three transistors, first of them wired as a colpitts oscillator. A common 3.58MHz ceramic resonator is used to produce oscillation. A varactor type tuning method is employed. I use a 5mm red LED as a varactor diode. Alternatively 1N5808 type diode can be used for this purpose as they exhibit linear tuning. A good quality ten turn potentiometer is recommended for the tuning control, preferably mounted with a calibrated rotary dial.

Q2 and Q3 are wired as a buffer amplifier that provides a reasonable output of about +17dBm. L2 should be wired on an eight mm dia slug tuned former with about 13 turns of primary and five turns of secondary. I used 28 SWG copper wire for the windings. If you use a toroid core for L2, add a trimmer capacitor parallel to C6 for fine tuning. C1 and C2 should be NP0 type or you can try styroflex type capacitors. L1 should be a moulded RF choke of about 10uH or more. In my version I just used the primary winding of an old 455KHz IF transformer with its internal capacitor removed. It measures 33uH. The VXO provides exceptionally stable signal even from the cold start. It provides about 80KHz tuning on 80 meters and about double on the forty.

2. DIGITAL VXO:  VXOs can also be built using digital invertor chips as these are cheaply available, these days. I have built many of them using 74HC04, 74HCU04 and CD4069 etc. The "U" suffix represents unbuffered version of the invertor and is usually preferred version for such applications. Though these chips are designed to be used at 5 Volts supply but they can work happily till 6 Volts and deliver a little more output; about 24mW. This level is sufficient to drive a diode ring type double balanced mixer. The schematic diagram of such a VXO is shown below:

C4 is an old ex-BC receiver component that can be replaced with a varactor arrangement for tuning, if required. L2 is wound using 32SWG, 11 turns close wound on a former made from the body of an old ball pen. The output low pass filter is included to filter out harmonic products that can cause erroneous harmonic mixing in the mixer, since being square wave the output of this VXO is rich in harmonics.

Both the VXO circuits illustrated above can be used in a variety of home brew direct conversion designs. Apart from ceramic resonators, three to four crystals can be used in parallel. I am using four 5MHz crystals cheaply available for use in microcontroller projects, in a bipolar VXO circuit mentioned above to provide coverage of QRP segment of 40 meter band on bitx. Mouser also lists 4.915 MHz ceramic resonators those can be used for a bit wider band coverage. 

An FLL can be used for more frequency precision. If you like to add a simple frequency counter, you can try Fredy's (DJ3KK) SPRAT counter. It uses an eight pin u-controller 12F675, thus it is very small and cheap and announces frequency in Morse, through a small piezo buzzer.

Friday 26 May 2017

XENA- A Portable DSB/CW Transceiver

In my post of 24 th March, 2016 I shared a simple direct conversion receiver project built around a home-brewed mock I.C. This project grows out from that design. 'XENA' is a simple, portable DSB/CW transceiver for backpack use. As mentioned above this cute little project evolved around a single home made mock I.C. (made around components shown in a dotted square). It functions both as a transmitter and receiver mixer. For simplicity the design uses a ceramic resonator/crystal based dual band, super VXO for 80 and 40 meters (more about this later). The receiver is developed around ubiquitous parts generally available in one's collection and the complete project is developed in modular form with all modules made using ugly construction technique. The modular construction allows scope for future experimentation and further development of the project. The schematic of exciter cum receive module is given below:

Even to a crass view the circuit of exciter cum receiver module is too simple and hardly needs an explanation. I have added a multi-turn preset in the biasing arrangement of the mock mixer I.C. to aid precise mixer balance in order to achieve minimum carrier leakage during transmission. All broad band RF transformers are wound on pig-nose balun cores using 30 SWG enamelled copper wire.
The RF amplifier circuit is quite self explanatory. The schematic diagram of RF final is given below:

The three stage amplifier ensures almost seven watts of RF output. I used 2N4427 for the driver as it was available in my collection but many other suitable candidates like 2N3866, 2N5109 and 2SC1175 seem to work as well. The receiver band peak capacitor C21 is a 330pF type variable tuning capacitor. It eliminates the need of mechanical band switching arrangement in receiver front end using switches or relays etc. An attenuator ahead of it has been included for dire reception phases. It can be switched in during presence of receiver overloading.

The entire project is developed on 4"X6" PC clad pieces using ugly construction and each board is fixed to the sides of the enclosure body. The control circuit being in the middle. The control circuit provides the required QSK delay during CW operation, does all switching and provides side-tone during CW transmissions. I used a small piezoelectric buzzer, the type used in computer motherboards, UPS's and microwave ovens etc. to provide CW side tone, since the idea was simple to implement. The schematic for control board is given below:

Sunday 2 April 2017

ZERON -A Super Simple QRP Dual Band Multimode Transceiver. -II

Just back after the hiatus, mainly due to a complicated leg fracture. However in this brief post I will share the direct conversion receiver part of "Zeron".

As I already told you about my obsession for digital chips for use in RF circuits and especially in switching mixers etc. I used an 'HC4053 as receiver mixer. This chip is an excellent choice over expensive diode mixers for many reasons. First as I already mentioned that it is cheap and easily available. The measured insertion loss is just about 0.5dB and the off state RF isolation is better than 45dB. The measured return loss is also better than 22dB for a variety of chips put under test, from different manufacturers. More than that you don't need to balance or match any devices! It is wonderfully simple to use.

The mixer is preceded by a pre-selector stage employing a common JFET in common source mode. For simplicity I used the air core coils but you can replace them by appropriate replacements in case compactness is more desirable. The variable gang capacitor is an ex- broadcast receiver type that facilitates the receiver peaking on the desired band of operation.

The AF component from the mixer is routed to a two stage AF preamplifier, through a duplexer circuit. The duplexer ensures that the mixer must see a fixed terminating impedance of fifty ohms on its output port, for an entire gamut of frequencies, literally from DC to daylight. The two stage preamplifier is modelled as constant impedance amplifiers, built around low noise transistors. 

A passive AF filter is placed after the first AF amplifier to shape the overall frequency response of the receiver. And as you can see it is the best place to place the filter in AF chain. The main reason of choosing a passive type of filter is that it has more dynamic range than its active counterparts. The inductors used are miniature encapsulated ferrite shielded units, generally available from Digi-Key. The two stage preamplifier is designed to eradicate any trace of fifty Hertz hum, generally considered a menace in most DC receiver designs. The two stage preamplifier provides about 60dB of gain. The following bode plot displays its frequency Vs gain characteristics:

I have built several receivers in past four decades ranging from re-gens, reflex, spontaflex, pentaflex, superjets and many others. I was delighted to build some and disappointed to hear others. One common last thing that can mar the overall performance of an otherwise good receiver design is; of course the final AF amplifier design. Consequently, I opted for a distortion free class-A amplifier employing a very low noise op-amplifier as its major gain block that supplies around 44dB of AF gain. The output of this amplifier is about 100mw and can drive a low impedance small speaker or a set of headphones.

The overall receiver performance is amazingly good. It sounded so well on crowded bands. There was no trace of overloading or broadcast breakthrough on forty meters but on eighty I had to switch in the attenuator on some occasions. The receiver performed as expected and was pleasingly sweet on ears.