By Shawn Baris ZR1EV
PO BOX 212
Brackenfell
7561
Republic of South Africa

*Anyone is free to use the following information for private use on the provision that it is not used for commercial purposes. Permission is granted by the author, ZR1EV, to publish and or distribute all or part of the following on the condition that recognition is given to the author.In the case of it being used in a newsletter or magazine, a copy of the aforementioned should be sent to the above adress.*

Having allways had an interest in "band 1" propagation, and being spurred by reports of TV signals from the old Crystal Palace transmitter on or around 45Mc being recieved regularly down here in Cape Town during the Seventies, It was particularly frustrating not to be able to operate on the 50Mc band. This band was for exclusive use by ZS-call holders and as you can see by my call, I hold a "VHF-only, no-code" ZR-call.

Things changed in 1981, when it was announced that the six metre band will be opened up for holders of the ZR-call on the 1st of January 1982. I was ecstatic! But elation was soon followed by disapointment when it became apparent that there were very few rigs to be had, and the prices of those available were way above what I could afford.

I was given an ex-military rig by OM Dave, ZS1SG, but soon realised that if I was to really exploit the amazing fare of different propagation modes possible on six metres, I would have to use SSB/CW and have more power at hand, when it was needed. A transverter seemed to be the obvious choice at the time, and I proceeded to aquire an SSB/AM CB rig, with the idea to convert it for operation on the 10 metre band, which I would then use to drive a transverter I was still going to construct.

After the mod to the rig, (which worked extremely well), I proceeded to experiment to find the best circuit for a transmit converter that would eventually form part of my prototype. At this point, I realised that the whole idea is not going to be all that simple. The local oscillator would have to have a frequency of 22 Mc. (22 + 28 = 50). Unwanted products from the mixer would be (28 x 2 = 56) and (22 x 2 = 44) These unwanted products from the mixer i.e. 56- and 44 Mc are only 6 Mc away from the wanted output and were going to be very difficult to attenuate to the required -50dB (min) level required and to have a consistant clean output across 50-52 Mc would be asking quite a lot from the mixer and filter stages. To cover from 50-54 would paint an even bleaker picture and would involve switching bandpass filters to enable the unit to stay within minimum spec.

At this point, something else dawned on me. For the rig to operate on 28 Mc, the local oscillator would have to be on (28.000 + 10.695 = 38.695).

The 10.695 is the IF frequency used on the rig I had and the local oscillator is injected on the "high side" of the desired operating frequency. It now becomes clear that if the Local oscillator signal could be treated as if it would be injecting on the "low side" of the desired operating frequency, the new frequency (image of 28.000, with 10.695 IF) would be 38.695 (LO) + 10.695(IF) = 49.390!

If the Local oscillator signal could now be made to go up to 39.305 (rig operating on 28.610) we now have (39.305 + 10.695 = 50.000!). This forms the basis of the modifications to follow. I part2, I will discuss the frequency synthesizer, and how to actually go about moving the rig up to round 28.600, which will be the first step in the mod. Also, bypassing the inherent problems when using a 40 CB channel switch in any mod, albeit to 10M or to 6M. (5 "gaps" where the frequency "jumps" 20kHz and the sequencing problems round ch23-25, and the nuisance of having to have a typed piece of paper stuck to the rig to tell you what channel=what frequency) The conversion is well worth the effort and I have already done dozens of these mods. The bugs in the early prototypes have been fixed and due to presure from the local guys, I have decided to "put pen to paper" and share what I found from years of experimenting and perfecting.

Hope you will be able to enjoy it as much as I have. Please let me have your comments and input.

For the purposes of this discussion, I will assume the following :

(1)That you have a working knowlege of RF circuitry and techniques.

(2)That you have the essential tools to do such a conversion.

(3) Basic RF test equipment.

(4) The set you plan to convert is fully functional.

(5)The set is already fitted with a 40-channel selector switch.

(6)That you have the service / workshop manual of your rig.

The chassis I will be discussing is the UNIDEN PTBM048AOX -chassis fitted to a number of CB rigs parading under different brand names. It is a single conversion radio on SSB and double conversion on AM, based on the popular PLL02A PLL chip, and has the Mitsubishi 2SC2166 and 2SC1969 in the TX lineup. Some versions of this rig operated in the "C" band (29MHz Marine) and had a diode matrix, six position channel switch, and could be set to step in 12.5 kHz increments. In addition, these units also had a "piggy-back" pcb fitted near the VCO, containing a 12.800 MHz oscillator. These sets should be changed back to the original 27 Mhz configuration andtested before any modifications are attempted.

The PLL02A(G) chip is quite remarkable in that it contains practically all the components needed for a PLL synthesizer, exept the VCO, of course. The chip sports a 9-bit binary divide-by-N-counter, a phase comparator with lock detect o/p (pin6), and a fixed divider of either "divide-by-1024 or divide-by-2048", selectable at pin 4. In this application, a 10.240 Mc oscillator feeds into this fixed divider at pin3, and when pin4 is left o/c (logic1), will divide this signal by 1024, resulting in 10kc, which will be your synthesizer "step". (If pin4 is tied to ground -logic0- the fixed divider will now divide by 2048, giving you a 5kc "step") DO NOT change the 10.240 frequency. The output of that oscillator is also used in the AM reciever section for the conversion of the 10.695 IF down to 455kc. (10.695 - 10.240 = 0.455Mc or 455kc) !

At first glance, it seems as if the chip might be able to cover over 5MHz in one go (512 binary combinations at 10kc cannel spacing),but when one has a good look at the spec sheet, the maximum frequency allowed into the divide-by-N counter is limited to 3.5 MHz. So, at 10kc channel spacing, the divide-N binary number of about 350 should not be exceeded to ensure that the chip runs within it's specs. This is the main reason that the mixer xtal (X1) 10.0525 MHz has to be changed.

IC2 functions as both the VCO and as a mixer (C3001/TA7310P). The interesting thing about this mixer is that it has two inputs (f1 and f2), and two outputs (f1+f2 and f1-f2). The difference output is taken off pin9 and the sum output is taken from pin6. The difference output is fed into the divide-by-N counter input, and the sum output is filtered and is used as the Local Oscillator signal. X1 (10.0525) and Q3 make up the oscillator and doubler stage.

As an example, channel 19 (27.185), the VCO will run at 17.775, mixed by the output of the doubler 20.105 by IC2 and will produce two outputs : the sum output, 37.880 and the difference output, 2.330. The 2.330 goes into the divide-by-N counter, where it is divided according to the value in binary applied to pin7 (MSD) to pin15 (LSD), in this case a value of divide by 233. The result of this division is 10kc, where it gets "compared" to the 10kc "reference" frequency from the divide-by-1024 fixed divider chain by the Phase comparator. The output from this "charge pump" is fed to a LPF where the "error" is filtered to DC, so that it can be used to either force the VCO frequency up or down untill there is no error.

The loop is now "locked" and pin6 will go to logic1 (5.4V). The PLL (Phase Locked Loop).

In part2, we had a brief look at the basic operation of the PLL frequency synthesizer (If you want a more detailed description, do drop me a line) and from that we can see how the channel frequency is generated. If now, for example, we change the binary input to the PLL chip to something else, say 234, it would result in the phase detector generating an "error", which will drive the VCO to the new frequency so that the loop will once again be in the locked state. As the VCO will now have moved 10kc, the resulting sum output will also have incremented by te same amount, resulting in the transceiver effectively having changed frequency by 10kc.

This scheme is fine for changeing frequency round the design operating frequency of the radio, but to change "band" to a frequency more than 1MHz up, one has to watch out not to exceed the 3.5MHz maximum frequency input to the divide-by-N counter. It should be clear that the synthesizer frequency can be moved higher by changeing the 10.0525 Xtal to a new value.

There should be no need to rush out and buy one, just have a look in your "junkbox" for a Xtal with a fundamental frequency round 10.400. After you have done the replacement the alignment is easy. Attach an RF voltmeter /'scope / RF "sniffer" to test point3 (TP3) and, using the propper (!) trimming tool, adjust the core of T3 for max deflection. Then, connect a DC voltmeter to pin6 of the PLL02A chip (take care not to short any of the pins) and carefully adjust the core of the VCO "block" until you get a solid 5V in that pin. Flick the channel switch from CH1-CH40 while checking the voltage on pin6 of the PLL ic and confirm that it stays 5V.

The PLL is now locked. Connect the RF Voltmeter/'scope/sniffer to pin4 of IC3 (C3001/TA7310P) and adjust T1 and T2 for maximum deflection on Channel20. Remember that the rig will not transmit if the PLL is not locked. When pin6 of the PLL chip falls below 5V, it "switches off" Q7 in the transmitter predriver stage, rendering the transmitter inoperative.

Set the rig in the transmit mode and realign the set as per the manual for maximum output. Rig should now be transmitting somewhere around 28.600 and the exact frequency can be checked with a frequency counter.(Please remember to have the rig in the AM mode for measuring the frequency - sounds obvious, but easy to forget in the heat of the moment)

You will now most probably be on a somewhat "wierd" frequency like 28.7842 or suchlike. (At no time touch the 10.692MHz oscillator trimming caps, CT4 and CT5 - I will explain their function later. They should not be touched at all if your rig was operating 100 percent on 27Mc). The easiest way to go about the following procedure, is to have an HF reciever on hand with SSB fascilities. In the previous example, say 28.7842, tune your reciever to either 28.780 or 28.790(preffered). Key up the rig again (in AM mode) and adjust CT1 for a zero beat on your reciever (Which should be set to USB) The rig should be keyed into a "Dummy load".

Now set your reciever to LSB and do likewise on your rig and while modulating, adjust CT2 for a natural sounding voice on your reciever. If necissary, experiment with the values of C20 and C21 to get the rig spot on frequency. You should get no less than 4W on AM and 12W on SSB after the mod, and if you want, you can realign the reciever strip now as well using an adjustable signal generator and use the set as a 10M rig, or get

ready to continue the mod up to six metres. Good luck.

Now that the rig has been converted to 10 meters and it has been thoroughly tested there, we can proceed with the conversion to 50Mc. You may by now have noticed that the chassis of the rig is isolated from the earth-plane on the printed circuit board itself. This was done to enable the rig to be installed in either a positive- or negative earth vehicle. The chassis must be at RF earth though, as the antenna socket must "see" the earth on the PCB. This is where a lot of "conversions" fall short in performance. So what needs to be done first is to replace all the disk ceramic capacitors that connect earth and ground together. All round the PCB where it is screwed to the metal chassis, you will see them. Don't forget the two caps on the DC connector ! Change all the capacitors with a value of 0.047æF (47nF) to 0.022æF (22nF).

Now for the reciever section. Change C100 (33pF) to 18pF and remove T7. Carefully disassemble T7 so you can get to the windings. Hold the coil former upright (pins facing down and the side with the 3 pins facing you).

CAREFULLY liberate the wire from the righthand pin and unwind the existing windings, taking care not to break the delicate strand. You'll see that the former has 4 "ribs" to hold the winding. Leave the wire attached to the leftmost pin and proceed to first wind 4 turns onto the second "rib" from the top in a clockwise direction. The "ribs" have a gap in the middle so cross over to the topmost "rib" and wind 4 more turns on there in the

same direction, feed the wire down to the righthand pin and solder it onto the pin. The secondary winding stays as is. Carefully reassemble, not forgetting the green plastic spacer on top of the coil former, and solder back into the printed circuit board. T7 is a Mitsumi type ETR0333.

Next, change C102 (47nF) to 22nF disk ceramic. Remove T8 and dismantle as with T7 and remove the small tubular ceramic cap. I found that a small surgical knife to be ideal here, as you can then just cut the leads of the capacitor, instead of battling to get them off the pins in an orderly fashion. You will now notice that this time the winding has a tap that goes to the centre pin. In this rig the tap is not used, so we can just rewind as for T7, except this time it will be 1turn second rib from the top and 4turns on the topmost rib back to the righthand pin and solder. Assemble and solder back into the printed circuit board. This time you have to add a capacitor to the print side of the board. T8 has 5 pins, grouped 2 on the one side and 3 on the other : the side with the 3 pins is the primary in this case. Solder a 39pF ceramic capacitor accross the outermost 2 pins of the primary winding on the foil side of the board. T8 is a Mitsumi type 10CA006.

Now remove T9 and strip down as before. This coil has a pink pot core and is a Mitsumi type 10CB001. Wind 3turns onto the second rib from the top and 1turn on the topmost rib. Solder back into the printed circuit board and add a 39pF capacitor onto the primary as for T8.

If your rig was operating for eg. on 28.790 on a specific channel, switch it back to that channel and calculate the new six metre frequency as follows : [10metre frequency] + 21.39 = [six metre frequency]. So in our example it would be : 28.79 + 21.39 = 50.180. Set your RF generator to that frequency and (using the correct trimming tool) adjust the cores of T7 T8 and T9 for maximum "S"-meter reading. Do not be concerned at this point

if the recieve sensitivety is not too good. We still need to change the Low Pass Filter on the transmitter section when we get to that stage.

You will also notice that USB and LSB have swapped around and dont seem to be on frequency... more on that in part 5

It is not my intention to provide lengthy and detailed explinations on the intricacies of SSB or bore those who know this subject well, but to point out some of the basics with particular refference to this particular chassis we are concerned with, and hopefully arm you with the neccissary basic understanding of what we are doing. This will be invaluable when fault finding needs to be undertaken at any stage.

I mentioned before that "USB and LSB have swapped around and dont seem to be on frequency ". To understand what has happened here we have to have a brief look at how SSB gets generated here. The "classic" method of generating SSB involves a carrier generator spot on the IF frequency feeding a balanced modulator of sorts. The resulting DSB signal is fed to switched crystal filters : one for USB, one for LSB, possibly one for CW and one for AM. As you can deduce from this, the manufacturing cost of this approach would result in quite high prices on the finished product being passed on to the consumer. This is probably one of the reasons that most dedicated Amateur Transcievers have the CW filter fitted as an option, rather than a standard fixture, in an attempt to keep the final cost to the consumer low and by implication competitive.

The designers of this (and other rigs) have cleverly sidestepped these problems with some ingeniuety. How AM is generated should be quite obvious so we will not dwell on that now but rather just focus on SSB. Inspect the innards of you rig. You will find only one large crystal filter located roughly in the middle of the PCB. This filter is about 3-2.7 kc wide and is a "LSB" filter centered round 10.6935. (Some rigs are fitted with a smaller rectangular filter, taking up about half the space. These are "metal-oxide" filters, but for all intents and purposes can be treated the same).

Suppose our carrier generator is set to 10.695 (standard 1st IF frequency). The 10.695 signal will pass through a balanced modulator chip, IC4 (AN612), where it is modulated by the audio signal in such a way that the original 10.695 is balanced out and you are left with a double sideband supressed carrier signal centered around 10.695. This signal is now fed into the crystal filter where the USB signal is heavily attenuated and them LSB part of the signal is only slightly attenuated (filter insertion loss).

We are now left with a LSB SSB signal centred on 10.695.What if we want an USB signal ? The answer is to lower the carrier frequency to below the filter passband : 10.692 is about right. The DSB signal will now be generated centered around 10.692, and after passing through the Xtal filter, will only have the USB components remaining.

Note though that the USB signal will be centered around 10.692 and the LSB and AM signals will be centered around 10.695. Voila ! we generated both USB and LSB with the same filter !

Now you can see that if the frequency from the synthesizer stays constant between "mode" switching, The LSB and AM signals will appear on the band "on freqency", while the USB will be 3kc "off". The designers got round this "problem" by shifting the synthesizer output frequency by 3kc when the USB mode is selected, thereby making all of this transparrent to the final user, and all the "channels" RIT exactly at "12'o'clock irrespective of the mode you have selected.

More to follow on this in Part6

As probably expected, the story does not end there. If you look at the circuit diagram, it becomes clear that both the frequency synthesizer and the carrier oscillator gets pulled LOW on LSB, not the other way round as implied in Part5 of this series. How could this be ? The answer lies in the mechanics of mixing. On 27Mc an inversion of the sidebands take place in the process of mixing the 10.695 DSB signal with the output from the frequency synthesizer (coming in from the "high side" of the wanted signal). The result of the mixing process and the resulting inversion of the sidebands means that one has to generate a LSB signal at the IF frequency to realise an USB signal on 27Mc, Hence the "pulling" low of both signals on both LSB and AM.

When we now convert the set to 50Mc, an additive mixing scheme is employed, and the sidebands come out the "right way round", with the result that the frequency shifting scheme would have to be reversed. In practice, this is quite a simple matter to accomplish. Some later articles on conversions like these in the Amateur press, circa 1985/6, would have you believe that it is simply a matter of redoing the labels on the front panel. This will not work and you will have great difficulty in getting the rig lined up to be on frequency on all the modes (if at all). Do the following and all will be well. You will see a grey wire running from a point marked (T) next to Q4 in the synthesizer section to another point also marked (T) on the pcb, next to Q18. Desolder this wire from the PCB where it enters the board next to Q18. Pull it through toward the mode switch. Next we need to look at the switch itself. There are two sections to this switch. The one section applies full B+ to the final when the rig is either in USB or LSB mode and routes the B+ via a hefty 6 ohm

wirewound variable resistor to the final when AM is selected (AM carrier power adjustment). Leave this as is, and locate the second section of the switch which selects USB/LSB/AM. The wiring from this switch goes to points (30), (19), (21) and (18) on the main PCB respectively.

Point (30) selects AM, point (19) selects USB, point (21) selects LSB originally and point (18) is common. Swap the wires from points (19) and (21) on the switch itself. Take two small silicon diodes (1N4148 or eq.) and tie their cathodes together and solder the free end of the grey wire to this junction. Shrinkwrap or slide an insulating sleeve over the exposed sections of this connection. The two "free" ends of the two diodes (anodes) solder directly onto the tags of the mode switch now connected to point (19) and (30) respectively. For those who are not sure, the end of the diode with the thickest colour band will be the cathode (apologies to those who do know HI).

The modifications to the mode selector is now all done, and you can now proceed to "net" the transciever. Adjust CT1 to net LSB and AM (one adjustment for both modes) and CT2 for USB only. To enable the rig to shift frequency on transmit as well as recieve, using the clarifier as a RIT/TIT controll, remove D5, and connect the unused end of the clarifier controll to the emitter terminal of Q44. Be sure to check the settings of CT1 and CT2 after this mod as it will affect the 12'o'clock position on the clarifier. I will discuss mods to the frequency selecting circuitry in part7... see you then.

In my introduction to this series, in part1, I mentioned the channel selector and touched on some of the difficulties encountered when using switches like these. This rig was designed for the American market and as a result are originally fitted with a switch that selects frequency according to the American 11M bandplan. This poses a few problems when any modifications are attempted to change the operating frequency.

For example, you will notice that there are five spots between channels where the frequency increments 20kc as the channel is incremented. Also, the frequency of channel 23 to 25 are not in sequence. Traditionally, mods involving a band change (say, to 10M) would require a piece of paper with a list of channel numbers with corrosponding frequencies typed on it attached to the rig, so one would be able to tell what frequency you are operating on.

In my mods, I inserted an EPROM between the channel selector and the PLL chip to "correct" all of these anomalies. It makes for a tidy, professional "feel" to the mod and makes checking your frequency as simple as looking at the LED readout. As the rig tunes in 10kc increments, one could program the eprom so that Channel 1 (readout 01) will put you on 50.010, channel 10 on 50.100, and channel 40 on 50.400 etc.

This is how to go about it :

The output from the channel selector switch will be a 6-bit binary "word" that will be unique for each of the 40 channels. We will use each of these unique 6-bit words to adress a specific location in an EPROM where you can program the 8-bit code of your choice. The EPROM will output this 8-bit word to the PLL chip and, voila ! you are on your frequency.

First you need to draw up a truth table. Make a list of all the channel numbers and write down the binary from the channel selector down next to it (you will find it listed in your service manual, or you can read the value by measuring each bit directly off the PLL chip. The pins in question are pin 15 (LSD) down to pin 10 (MSD), in that sequence). Next, convert the binary words to HEX. All EPROM programmers I have seen work with HEX.

Select any channel above 27 (ch26-40 are all in sequence). Read the full 9-bit binary off pins15-7 and write it down. Write down the frequency the rig is operating on next to it. Click one channel up and do the same. The binary should have changed by 1. Calculate all the binary codes for the range 50.010 up to 50.400 and when that is done convert them all to HEX. You now have all the information you need to program your EPROM. The eprom has an 8-bit output and the synthesizer chip requires a 9-bit word. In practice you will find that the state of the MSD pin stays the same throughout the 40 channels, with the result that pin7 can be wired permanently to either logic1 or 0.

I use a single 27C64 eprom, as there are a lot of them floating around surplus. Remember to "pull down" the first six imputs with 1k resistors and tie all the unused inputs to GND. I used a seperate 7805 3-terminal voltage regulator IC to power te eprom and all these components can be mounted on a small pcb or perfboard. I used short sections of ribbon cable to carry the data from the switch and to the PLL chip. It is of course possible to extend the basic 400kc tuning by programming another set of codes for say,50.400 to 50.800 and then use one of the unused inputs to the EPROM to switch "bands" this way.

See you in Part8 where we will start the mods to the transmitter.

Whenever any modifications or repairs are done to any piece of equipment, capable of radiating RF energy, it is of paramount importance that the regulations concerning spurious and harmonic energy radiation are strictly adhered to. This neccesitates the use of a spectrum analizer to verify signal purity after these mods are done. I realize that a piece of test equipment in that class is not standard equipment in most "shacks", but one would be able to most probably approach your local PMR company to check the unit for you to ensure that your output is "clean" before you transmit into your antenna.

The output from the frequency synthesizer is fed via a capacitive divider circuit to pin4 of IC3, where the IF injection frequency (injected into pin1) is mixed with it to produce the 50Mc wanted output. There are what appears to be two small chokes feeding pin6 and pin8. L23 is indeed an RF choke, but L3 is not.If you study the circuit carefully, it will become clear that C37 and C38 are in series and connected across L3 to form a parallel L-C Tuned circuit, which at the moment resonates at round 27Mc.

C37 and C38 also form a capacitive divider to do the impedance transformation to the imput impedance of the cascode amplifier section of IC3 (input on pin7). L3 is wound on a carbon resistor of round 400 ohms to lower the Q of this tuned circuit and thus improve stability and to increase the effective bandwidth.

Initially, I used to just take a few turns off the coil, but I have since decided to make a whole new coil instead and just replace the existing L3 with the new one. I used a slightly thicker wire, and wound exactly 15 turns onto a 470 ohm 0.5W carbon resistor. Cover the assembly with heat-shrink tubing, and apply moderate heat till the shrink-tube contracts and holds the windings in place. Remove C37 and C38, and insert 47pF ceramic capacitor in C37 and a 27pF ceramic capacitor in C38. I have found that changeing L23 there was no percievable improvement, so, youcan leave L23 as is.

The output of the cascode buffer amplifier is fed into T4 and T5 which forms a bandpass filter. Remove T4 and T5 and dismantle as described in Part4. T4 is a Mutsumi coil, part number 10CA006 and T5 is a Mitsumi, part number 10CB003. Refer to part4 as to the orientation of the coils. T4: Wind 1 turn onto the secon rib from the top starting at the leftmost pin. Feed the wire down the gap provided on the former down to the centre pin, form a pigtail, tin, and solder to the middle pin. Feed the wire back up through the gap in the former and wind an additional 4 turns on the topmost rib of the former. Feed the wire back down the gap, tin and solder onto the righthand pin. Remove the internal capacitor, reassemble and solder back into the PCB. Solder a 39pF iniature disk ceramic capacitor across the primary (Ref.Part4)on the foil side of the PCB. T5:Starting at the leftmost pin, wind 1 turn onto the second rib from the top, crossover at the gap and continue to wind an additional 4 turns onto the topmost rib, feed the wire down through the gap and tin, solder onto the righthand pin. Remove the internal capacitor, replace the black potcore, reassemble and solder back into the PCB. Add a 27 pF miniature disk ceramic capacitor to the primary on the foil side of the PCB. All the secondary windings can stay as is. Remove C42 and replace with a 2.2pF ceramic capacitor. Remove C43 and replace with a 39pF. Remove C185 and replace with a 18pf Ceramic disk capacitor.

We will continue with the modifications to the transmitter in part9

We are now done with the "encapsulated" small coil formers and the rest of the work can now be done without the aid of a magnifying glass (HI). Locate and remove T6. This tuned interstage coupling transformer is wound on a white plastic former and a lot of care should be excercised not to overheat any of the pins, as the former may deform from the heat and render it useless. The tuned primary of the transformer, can easily be identified. The primary coil is wound from clear enamel coated copper wire and the 1 turn secondary winding is wound using green enamel coated copper wire.

Leaving the secondary winding as is, proceed to remove 2 turns from the "top" of the winding, untill you are left with 5 turns on the former.

Form the free end of the wire with a small "longnose" pliers and feed it through the hole in the base of the former where the original lead was fed through. You will now notice thet the secondary and primary windings are spaced too far apart (by the thickness of about two windings). "Pull" the secondary winding lower down on the former (by pulling in the secondary pins with a small pair of pliers) until the two windings touch. Trim the pins to length and using a small hobby knife, scrape the insulation off the pins and tin. Reinsert the transformer, solder back into the PCB and replace C47 with a 27pF Disk ceramic capacitor.

Replace C49 with a 120pF disk ceramic capacitor. Replace L6 with a coil consisting of 13 Turns wound onto a 1.5M ohm 0.5W carbon resistor. Replace C52 with a 47pF 50V disc ceramic capacitor and C53 with a 120pF 50V diskceramic capacitor. Remove L7 and remove turns until you have 1.5 turns on it. Prepare as with T6 and replace. Replace L9 with a coil made up as for L6. Modify L11 for 2.5 turns, L13 for 3.5 turns and L201 for 1.5 turns.

If you have difficulty locating L201, you will find it (airwound coil) between point (31) on the main chassis and the inner conductor of the antenna socket. (some rigs have L201 and C201 omitted). Change C51 to 22nF

Replace C54 with an 82pF 100V disc ceramic capacitor, C55 with a 330pF 100V capacitor, C184 with a 68pF 50V disc ceramic capacitor and C201 with an 18pF 100V disc ceramic capacitor. As "all coils are not wound equal", You could experiment with slight changes in capacitor values for maximum performance with minimum spurious output. C56 is fitted to some chassis, and can be removed in this mod. L13 is an air-spaced inductor with self supporting windings, and can be streched or compressed for maximum output When you notice that the peak tuning on T4 and T5 occurs when the slugs are about the middle of the formers, you could INCREASE the capacitor value across the primary by a few pF. The correct point for the cores are about flush with the top of the screening cans. The same applies for the Reciever coils, T8 and T9 with the exception of T7. T7 should be tuned for the loudest signal, not the loudest "hiss". This is important for a good s/n ratio, so you can dig out those really weak signals on the band.

I have found that this occurs when the slug of T7 is a couple of turns from the "bottom" of the coil former. If you cannot get this condition, try changing the value of C100 by a couple of pF.

The rig should now "run up" smoothly. All of the rigs I have done so far, exceeded the original specs on 27Mc on the reciever and practically always got in excess of 10W PEP out of the transmitter (recomended PO). If you have a "lazy" 2SC1969 (Did a conversion once where I could do what I liked, but could not get more than 8W PEP... Replacing the final did the trick), You could try replacing the final with a 2SC1307. These devices seem to have more gain at 50Mc. See you in Part10 (final)

The conversion I have described in Part1-9 has been in dayly use in my shack for many years, and is the only rig I use on the six metre band. I have done countless conversions like this for other enthusiasts and we have all had hours of enjoyment out of them. The rig performs well on most of the modes of propagation found on the six metre band, and with the addition of a couple of extras, produces a stasion that you do not have to apologise for.

I built up a push-pull linear amplifier using 2 x 6146B valves and my rig drives them to max output easily. In addition I constructed a masthead GaSFET preamplifier for that extra edge during weak openings. The antenna will be one of your most important additions (as is the case with any station, on any band). It would be plain folly to expect "big" results from a "small" antenna.

For CW enthusiasts, the addition of an 800Hz audio phase shift oscillator feeding into the mic input would produce excellent results. I had a TONO 5000E connected to my rig for CW contacts with JA stations and had many QSO's on CW with compliments recieved on the "tone".

The reciever is sensitive enough for serious meteor scatter work, and had worked many stations on this mode. The rig lends itself well to mobile operation and I used to often sit in the early morning traffic chatting to

the locals on 50.200. You would be surprised at the excellent signals on 6M SSB when you are used to FM repeater operation.

Six metres is my favourite band and allways will be, and I hope you will share in the excitement and thrill in homebrewing your own rig and using it on your next DX contact. It was great fun for me and I hope it will be for you too. Finally, an extract from my log book, using this rig,to wet your appetite :

Happy mods and I will be on the lookout for you the next time the band opens !