Connect with me today

  • Call me on 44 (0)7970 190437

All sorts of interesting stuff about Ham Radio and old gear from Steve G3ZPS, near London in the UK

KW Electronics Ltd - A very British Company

Early in 2016 the nostalgia bug bit again after talking to quite a few chums who had connections with KW Electronics (Dartford UK) or who owned the kit. This Company had a special place in my teenage Ham years as my good friend Paul G4BXT (now sadly SK) lived near the works. On some Saturdays the two teenage hams would go into the showroom to play with the shiny new KW2000B's, KW Atlantas and the separates. I found out later that two friends I met in 80's worked for the Company before it folded in the 70's


KW started life in January 1956, when Roland (Rowley) Shears - G8KW and Ken Ellis -G5KW, began importing Geloso VFO's from Italy and making antennas, dummy loads and accessories for the UK ham market in Rowley's house.  This rare ID Plate shows Rowley's house address, its on a TVI filter to notch out UK VHF TV CH1 on 45MHz. Part of my collection and it dates from the late 50's

AM transmitters and receivers quickly followed with the  KW Valiant, KW Vanguard - and cute little KW160. SSB equipment appeared in the late 50s but in the 'boat anchor' style - the early Mk1 KW Viceroy SSB transmitter had a huge separate PSU.

KW even managed to get one of their AM Transmitters into a James Bond film - Dr No - all British line up, KW and Eddystone !

The KW2000 and the 'G LINE'

Early in the 60's Rowley and his small team must have realised that smaller (and lighter) SSB equipment was going to be the future of the hobby on HF. In the US  the Collins Radio Company had started the trend with their innovative SSB transceiver - the KWM-1, shortly followed by the legendary KWM -2 - a masterpiece of radio design. The KW team took the basic concept of the Collins equipment and re-designed it so it could be manufactured in the UK and sold far cheaper than any imported US equipment  - the team called it 'Project 2000' . This rare picture of the prototype looks really nice, although the radio become a bit cluttered with the addition of VOX and IRT in production. The original press cuttings at the time of release say that the radio was specifically designed with mobile operation in mind. They also indicate the radio was fitted with VOX, but the odd 'VOX BOX' in the MK1 2000s is not in any of the photos of the prototype.

Corners had to be cut however - the Collins permeability tuned VFO (PTO) was replaced with a conventional design - The front end and driver tuned circuits on the higher bands were wound directly around the ferrite slugs (and glued in place) and placed in a series / parallel arrangement with the 160/80m coil (makes alignment a bit more tricky!). The premix crystal oscillator coils are also wound the same way and this makes alignment very tedious - you have to have a bottle of nail varnish remover to hand !. Note also that the RX RF amplifier screen resistors are gradually switched out as you go up the bands. This is probably to try and overcome the losses introduced by the poor front end design on the higher bands and was not improved until the 2000E came out.  The Heathkit SB and HW line use the series parallel coil approach as well, and leave the RX RF amplifier at high gain. It may be unkind, but compared to the superbly engineered and documented Collins KWM 2/2A the KW2000 was clearly more of an 'amateur radio' product.

The basic radio architecture is the same - Dual conversion, broadband 2.955 to 3.155MHz first IF, 455kHz final IF, Kokusai Mechanical filter (instead of Collins), 200kHz band segments, 6146 output tubes, KW also added 160m,  IRT, VOX and the turned aluminium knobs added a lovely touch. Thus the very British KW2000 series of HF radio transceivers were born. The earliest model released in late 1963 had one 6146 and was very compact,  it had a tendency to drift if used mobile as the vehicle battery voltage and the VFO heater supply changed - pull away from the lights and move up the band a bit!. Despite these niggles it was far cheaper then the competition and sold well, although clearly some improvements were needed. They first came in the form of the KW2000A model introduced at the end of 1964 for shipping in 1965. Two 6146 PA tubes were fitted into a deeper chassis and the VFO and premix valve heaters were run from a separate 6V supply to cut down the drift with supply voltage changes. With the launch of the  KW2000A, the cases were redesigned to match across the expanding SSB range and KW called it the 'G line' - a clear copy of Collins 'S Line' philosphy. 

Despite the introduction of the KW2000A in early 1965 the original 90W MK1 stayed in production alongside it for a further 2 years. Although either rig could be used with a mobile DC supply, the smaller size and lower power consumption of the MK1 must have ensured enough sales.


Its quite difficult to trace the timeline of the early KW2000 radios, but as far as I can tell looking through the KW magazine adverts there were at least 4 distinct variants of the MK 1 as KW fiddled with the design ;

* The very first production run was quite short and ran from late 1963 until around May 1964. These models had the meter design from the prototype (used on the earlier boatanchor radios as well) and a very simple dial escutcheon (the prototype had no escutcheon). You can see that KW were clearly aiming at the emerging HF mobile market in this picture. These early variants are quite rare and I have not seen one for sale recently.

* The 2nd production run from May 1964 until  Summer 1965. These models had the meter and escutcheon that was used on the later 2000A but were not in a 'G line' case and had the light blue front panel.  The picture makes it appear as a slighly better looking home station ! I have 2 of these

* The final MK1 KW2000 production period ran alongside the 2000A in 1965/6 and had a small G line case (with the inspection lid and no mobile mounting bracket fixings). Early models of this period had the light blue front panel and later ones had the dark grey front panel (same as the 2000A). My friend Richard G3UGF believes there may have also been an option to buy a separate G Line case to improve the 'look' of the older models.I have one of these with the dark front panel.

The VFO dial on the 2000 / 2000A has a readout that just about manages a resolution of 2kHz and the odd (and unobtainable) slow motion drive has a tendency to wear so that you set a frequency and then the dial moves a little..called 'KW lumpy VFO syndrome'. There aren't enough 200kHz segments to cover 15m and 10m - another compromise. KW quickly added the 'separates' to the G line - the KW201 receiver and the 'Vespa' transmitter. Both of these were updated to become the KW202 RX and KW204 TX, but even into the 70's these lacked basic 'transceive' capability and had to be netted onto the same frequency. The KW 1000 linear amplifiers and antenna tuners have endured and many are still working in ham shacks to this day.

By 1969 the storm clouds were gathering for UK and US Ham Radio manufacturers, but KW managed to upgrade the 2000A to the 2000B and found a real winner in the UK market. As well as a few minor design tweaks the big improvement was a 2 speed dial mechanism with a slightly better readout and the rig sold like hot cakes for a few years. There are stories of queues of UK hams turning up at the Dartford Works on a Saturday to pay cash for their new KW2000B !. The last gasp for the 2000 was the 2000E from 1973 until around 1976 (its missing from the 1976/7 KW catalogue). Although it was the first (and only) complete redesign of the KW2000,  had an improved receiver front end and 500kHz band segments, it suffered from further cost cutting, strong competition and limited development. There was no denying that by the time of is introduction all valve HF SSB transceivers were a dying breed  (it turned out to be a slow and painful death for Drake, KW and Swan amongst others).  The commercial variant - the KW2000CAT - channelised radio telephone sold well into the 70's and in its last incarnation used solid state Plessey devices for the low level signal generation.

KW tried to break into the US ham market in 1969 with its SWAN 500 copy - the KW Atlanta (more on that further down this page). The Company was sold to Decca in 1974 (Decca were more interested in the profitable commercial arm of KW than developing the ham radio line), but by then sales of Yaesu, Icom and Kenwood had grown rapidly in the UK, and - shock horror! - they used extensive solid state circuitry. The Yaesu FT101 was the epitome of Japanese design and innovation - it was the next generation. The last catalogue of KW ham equipment is from 1977. Decca were bought by the Racal Group in 1980 for $250m - Racal were keen on Decca's expertise in microwave radio and electronic warfare systems, but most definately not in ham radio!. Rowley managaed to buy the rights to the  KW name back along with quite a lot of assets and carried on trading as a ham radio supplier selling KW ancillaries and TEN TEC rigs until he retired in 1989.

Peter Chadwick G3RZP worked for KW for a few years and wrote an excellent article in 2010 on the KW2000 series, I have linked it here.  The inspirational founder of KW - Rowley Shears died in 2009 - and only after his death did his amazing wartime radio exploits come to light - but that is another story!.  There is also an avid group of KW collectors and restorers on the KW Yahoo Group. Every January a group of KW enthusiasts get together on the HF bands and celebrate the founding of the Company in 1956..look out for us as GB8KW, GB2KW and even G8KW (now issued to to Rowley's son Richard)


I have had no real problems in getting any KW gear back on the air - the simplicity of the designs means most faults are easy to find. This valve gear is best looked after by those who really know what they are doing, you need to get inside and be aware of the danger of high voltages. I have seen some really daft questions on the forums like - ' my Drake TR4 has stopped transmitting..any ideas?' . All the valves used in KW kit are available, passive components are easy to replace and even the mechanical IF filters can be replaced with more modern Murata CFJ455K variants (although these are also getting harder to get).  Resistors go high and are easy to check, Hunts caps are awful (but not used in every KW), the resistor trimpots used in the 2000 RIT and S meter circuitry get dirty and cause problems (erratic S meter, different TX and RX frequencies) and I change them. Relays can be problematic, clean the plug in pins with metal polish and the contacts with non abrasive cleaner. Disc ceramic caps are pretty reliable. I have built up a stock of parts that should last me out. I only restore radios that are mostly complete, total wrecks are not for me !. The biggest enemy of any restoration project (radios, cars, houses!)  is time, commitment and the workshop facilities and spares to finish the job. Despite the good words of many hams with restoration projects sitting under the bench, I know very few will get those old radios back on the air. All the radios on these webpages are working.

EARLY MODEL 'PRE G LINE' KW2000 (X 2 !) MAY 2017

Took delivery of 2 early 2000s in May 2017. The serial numbers are close (299 and 313), probably 1964 vintage. Came from the same eBay purchase complete with Power Supplies. One of these radios (on the left) had been stripped down and partially restored, the other had some new parts fitted but very shabby. The'Pre G line' cases are very thin and poor quality, you can see why KW moved up to the better quality cases.

 There was a big difference in the state of these two radios, the one on the left had been very well cleaned up - someone had worked hard on it. With the unrestored KW2000 (right) I started to encounter some new issues...for some reason the ferrite slugs in some of the coils had been exposed to damp and had jammed rigid - what ever I tried (oil, wd40 ete etc) I could not move some of them. If was ever going to align the radio properly after restoration I had to do something. The only option was to remove the coils  and very carefully try and chip the core out. I ended up using a drill and a 6mm tap and took it very easy so that the formers were not damaged. the worst was one of the broad band coupler coils with has 2 slugs. This radio also had a lot of component changes and not all of them the correct value..ARGHHHHH, the bodgers..why dont they get it right or leave it alone?. At least he didn't use an electric drill on it! . The radio was extremely dirty with quite a few dry joints, wiggle a suspect component and it comes off in your hand. Surely no one used it with the very odd parasitc choke?.its just a coil of wire. As with my other projects I removed the valves, crystal, relays, vox, PA and the front panel so I could clean both radios up. These are just some of the issues I have faced fixing KW2000s..

  1. Wrong value or out of spec resistors on the screen grids and anodes of the RX RF amp and IF amps. Check all cathode resistors. Only a few electrolytics in the actual radio and only one on the HT

  2. ALL HUNTS black axial capacitors must go. No matter what you think about the radios performance, they are truuly awful . They were identified as a problem in TV and Radio sets as far back as the 60's with some estimating a design life of only 5 years- if you find them they must be replaced. They are mainly present in earlier KW gear.

  3. The rubber band holding ferrite slugs in the coils can turn to cement - they will never turn again. First time I have seen this.

  4. Broken central spigot on the 6146 final in one radio..and yes, I did plug it back in wrong as I hadn't noticed..lucky there was no bang, it just upset the heater balance

  5. VFO faults -  low output or only oscillates at the bottom of the band..if the resistors checkout OK (you can test a lot of them with the valve out and measure resistance between the valveholder and the 150v HT or ground), it may be nasty HUNTS bypass caps inside the VFO. I test the units with 6.3v and 150v on an external power supply and a scope. The difference in output with new caps is staggering. The VFO has to come out though ! This pic shows the location of the caps (one is hidden behind the resistor)...I have snipped the leads to tack on 2 new caps and check the VFO operation before the final fix.

  6. My NOS resistors may only last 30mins before failing.

  7. RF choke in the cathode of the 2nd TX mixer OC I know why I had no power out. Also checked the RF Chokes in the Driver stages, often burnt but still sort of working. Replace

  8. Always replace the PA grid coupling cap and check all PA components for signs of stress

  9. More dry joints than I had seen in a while (in both of them). This was mainly due to botched component changes and mods rather than KW assembly faults...and the odd Hunts black capacitors that had been left were changed.

  10. The strange purple wire in the wiring loom near the Mic socket of Mk1 2000s...doesn't go anywhere..checked the other radios..same wire just with a bit of sleeving over the end...odd

  11. Very poor carrier suppression  - more HUNTS capacitors, leaky diodes (only 2 of them) can be replaced with modern BAT85. One of the 453 or 457kHz carrier cystals has drifted into the filter passband or vice versa!.

  12. Strange AGC performance will probabaly be due to poor / leaky RX IF amp valves (both 6BA6), or the EF183 RF amplifier.

  13. The PA driver 6CH6 can take off without its screening can...same as Heathkits !

  14. The rear 2 bandswitch sections wear and mean the PA coil taps may not connect reliably. Its not so much the contacts, the rotor part is worn where the shaft goes through..I think replacement is the only solution - but difficult 

  15. Complete realignment (premix crystal oscillator is often way out on a least one band), and dont forget that nail varnish remover to soften the glue on many of the coils wound directly on the ferrite will never ever align the radio properly without it !...the premix crystal osc drive is critical for the TX, you wont notice it as much on receive. Put a scope on the EF91 crystal osc output and tune each band for maximum

  16. The mechanical filters can still work after 53 years!


OH DEAR !                                                    ANOTHER BODGE                                                NEW CAPS SLEEVED TO FIT PSU

After an extensive strip down, deep clean and renovation both Mk1 radios are working well on the low bands.  The PSUs also got a complete refurbishment, I replaced all the capacitors, resistors and associated components - the left and middle pics above show how bad things had got and how Mr Bodger fixed it (no capacitor mounting bracket). I replace all capacitors with modern small components, but make up proper fixing hardware (right). I reconditioned the paper cone loudspeakers and baffles and replaced all the tarnished screws (including new cabinet feet). My research on the case colour of these early radios threw up some odd results - very early models may have had a darkish blue case, although it appeared this quickly changed to a light hammered grey. The bezel is part of the case and cant be easily removed (although on one of mine the rivets had been drilled out of the PSU bezel so it could removed).  I rubbed down and test sprayed one of the cases with hammerite smooth silver which gave very pleasing results and was close to the original colour. Click on the pic above for a larger image of this fantastic restored radio. The microphone is a c1970 rebadged 'Turner + 2 CB/ Ham power mic', all metal made in japan and picked up at a rally for £10 - works fine.


This scruffy and rather shabby later period Mk1 looked very sad on first inspection. It is period correct with a single 6146 PA, in the small chassis, better 'G line' case and with the strange VOX BOX. This odd box is shoe horned into all the early 2000s on a large multiway connector in what looks like a real bodge job. Oddly there is an unused valve holder hole in the chassis under the VOX BOX of these radios - so it looks as if some last minute design changes were made!. Although KW called it an option the radio won't go into transmit without this unit. I think its a pity that they had this extra unit fitted , the prototype 2000 didn't appear to have it and looked very nice indeed. The MK1 2000 had an odd production run from late 1963. By December 1964 KW was advertising the improved KW2000A model but continued to market the MK1 until the end of 1966 (advert on left from August 1966) . The only explanation I can come up with is that KW stuck with the MK1 as its smaller size and lower output power made it better suited to mobile operators. It was not in any KW adverts by the start of 1967. This example is from the later period of production as the front panel is the darker colour of a 2000A

This 2000 came from a museum that had closed - it was bought untested for display and had not been used since sometime well before 1993.


I left the very dirty PSU to one side and checked the radio first. A thorough and deep clean normally reveals the the first few problems.  Nothing for it but to remove the front panel, the vox box, all the valves, relays, PA cover and get to work with soapy water and a stiff brush (not quite the soaking that some Collins and Heathkit restorers advocate). The aluminium knobs are particularly hard to clean up but can be brought back to reasonable condition. As expected some issues soon came to light. The PA Anode RF choke had been replaced - badly, the 6146 looked in poor health, the VFO slow motion drive had failed and a different Jackson one had been grafted on. The case was also partly repainted (?) in the wrong ham bodgers should be kept well away from spray paint!.


I changed the components that normally give problems (screen and anode resistors / bypass capacitors), plugged it in to a spare KW PSU and it came to life !. After a touch of alignment and a rebuild of the PA I dared to transmit..and after a little bit of adjustment we had good power out on 160, 80 and 40m. It drifts less than my 3 other KW2000s and by all accounts sounds fine on 80 and 40m, and how did the mechanical filter manage to last over 50 years?.

The 2000, 2000A and Vespa used a strange Jackson (?) slow mo drive and a dial cord to turn the VFO. - these drives are unobtainable in the 21st Century. Although Jackson drives are available again, to the best of my knowledge this type has not been remade with the precise dimensions to drop straight in. What's odd about it is that its just a concentric drive, the pulley sleeve is the output..there is no shaft at the rear. The most common failure mode is that the ball bearings fail of break up and the central shaft is often jammed. The VFO still tunes but really quickly !. I had to find a solution and had a spare 2000 VFO to practice on - I found I could use a new Jackson 4511DAF drive  (as used on the 2000B) with some small homemade aluminium brackets. It was fiddly to make them and I spent a few hours with the needle files until I was sure it would work. When I make another set for my Vespa drive I will make them a fraction and learn.

The dial cord only has a light load and works fine on the outer part of modern drive. I did not have to make any mechanical changes to the radio or the VFO to do the the mod can reversed if another drive turns up. A previous owner had already drilled extra holes in the front panel, my mod didn't need them -radio ham bodgers should be kept away from electric drills !.  Once released from the main chassis the VFO and drive can be removed as one complete unit. The hardest part was measuring and getting the dial cord on. I came up up with a work around for that as well - I pretensioned the dial cord spring with a thin pice of wire and got the dial cord as tight as I could, then with it all assembled I cut the wire holding the spring and it pulled the cord very tight..needs must. If you look carefully at the pics you can see the two little holes I drilled in the big VFO pulley to hold my wire tensioner..neat !!



The finished radio looks brand new - case repainted in 'Ford Polar Grey', bezel repainted in 'Ford Dove Grey' and the new slow motion drive now looks and works as it should. I re-capped the Power Supply (which did have Hunts capacitors fitted), I just dont trust 50 year old electrolytic caps. That also came up OK and after a repaint the pair look absolutely fantastic.

FEBRUARY 2017 - KW2000E

Never expected to find a rare super clean KW2000E, but I did. At one of my radio clubs a local ham heard me talking about KW equipment..he offered me his KW2000E at no cost as he was preparing to move house..would not take a cent for I made a charity donation for my conscience. I got lucky yet again with a radio that had hardly any marks and had been properly stored for over 20 years. Came complete with PSU and manuals. One day of cleaning pots, switches, valve holders and 'soft start' power up saw it come to life without a hitch.

Initial under chassis inspection revealed a new RS O/P transformer and some obvious resistor changes where the mechanical filter connects to the first RX IF amp (6BA6). I am reluctant to put all the values back to stock in case there are other mods that I have not noticed. The VFO stability is better than my KW2000B (the E takes less time to settle down). Uses a couple of rare tubes that I did not have in stock - 12BZ6 RF amplifier, and ECC85 first RX mixer.

Its a strange radio with a mixture of improvements, poor design choices and cost cutting over the earlier 'B' model - much better front end on RX, 500kHz band sections but the proper tube S meter amplifier of the A and B was left out (although a spare relay cutout can be used to put it back in ).  The rear octal connector for an external VFO (as on the 2000B) is also missing and blanked off. As the E has 500kHz band segments KW could no longer use the broadband first IF of earlier models and had to tune this IF in parallel with the VFO. This meant adding extra gangs to the VFO capacitor which would no longer fit inside the existing the capacitor sticks out the back of the now open VFO in what looks like a horrible bodge. KW had the same issue with the KW202 receiver.

KW never got to grips with a decent 1kHz dial readout either, the 2000E was a step back from the 2000B, and was only graduated every 5kHz. This is a legacy of using a Jackson slow motion drive, Drake and Yaesu sorted this years before by using a bespoke assembly for their VFOs. Quite why KW bothered with a WWV band position is also beyond has a 100kHz calibrator, and with 5kHz readout - its a pointless addition. I have also noticed some spurious signals on receive that are hard to pin down..odd as the mixing scheme is the same as other KWs and they don't have them.

The 'RF Gain' control is another odd part of the design. KW had a lot of overload problems (probably on 40m) on their earlier rigs and to get round it just stuck a 5K pot across the antenna where it goes into the not a gain control at all, just a crude variable front end attenuator (the KW202 Receiver has one). Unlike the AGC  gain control of the A and B models, on the 2000E there is no way to manually adjust the RF and IF gain of the receiver at all. Things got so bad that KW offered a 'modification kit' to replace the RF gain control of the 2000B with this front end attenuator. For some inexplicable reason KW never offered metering for either Relative Power Output or ALC even on the 2000E, all the Heathkit, Yaesu and Drake radios have these features..what were they thinking ?. Its a real pity that KW couldn't deliver with the E, it was probably conceived as a stop gap until the doomed digital readout KW2000D was developed (only prototypes ever made)... but unfortunately it was too little and too late and signaled the end of all development on KW ham radio gear.

I had trouble with one of the  2 12V 4 Pole Relays in this rig..didn't take long to spot part of the problem..when I removed the cover to clean the contacts the base of the relay had cracked so much it literally fell apart !. The culprit appears to be excessive heat as the relay is close to two very hot parts of the radio..the audio output valve on one side and a large wirewound resistor on the other. It was almost too hot to remove with bare fingers.  I had a spare but have since located a supplier of modern plug in replacements from UK supplier RAPID

Plenty of contacts on 40 and 80 but may need alignment on 20 / 15 / 10m

OCTOBER 2016 - KW 2000B

It took me many months to find a really clean KW 2000B, but one turned up on eBay in October 2016 just a few miles away. The key for me was that the front panel was unmarked  and it looked like it had been owned by a lab. It had BT (UK telephone) stickers on it and had not been bodged, abused or fact very few marks. It was bought at an auction a few months before as it looked nice, but its owner couldn't get a sound out of it and asked his mate to sell it. The PSU was also very clean and he said the radio lit up. The main cosmetic damage was to the radio bezel which had been covered with sticky tape to protect it - although it still had some marks. Year unknown...sometime between 1969 and 1973

Once home it was obvious in minutes why the radio made no sound when he plugged it in..The headphone socket had failed internally and left the loudspeaker permanently disconnected...fixed that it came back to life on RX.  All relays came out for cleaning of both the sockets and the actual contacts, bandswitch was cleaned and we held our breath for TX..again no problem and contacts made straight away. VFO drift is quite bad on this radio even after an hour warmup and I have bought a Cumbria Designs X lock 'Huff n Puff' VFO stabiliser to make this 2000B usable without constant tweaking of the VFO. This is an easy fit as the 'Cal' control can be used to feed the correction voltage into the VFO.

Fitting an X lock

The X Lock is now fitted to the KW2000B and has stopped the drift in its tracks. The X Lock needs to voltage control the VFO and there are a number of ways this can be achieved. In the 2000B I decided to use one of the varicap diodes already fitted inside the VFO to get the dial calibration correct. My initial implementation meant the disconnection of this 'calibrate' input to the VFO (purple wire) and connection of the X Lock directly to this pin. I then re-adjusted the VFO coil to bring the it back to the nearest 100kHz point. There is a slightly better option which I may move to that would allow the CAL control to have a small effect, this sums the 2 DC voltages together with a couple of opamps.  I have been careful to ensure the mod can be reversed without any changes to the radio..should a purist own it in the future. I powered the board from the 12V AC heaters with a single diode and cap to get 14V DC. Look at my video here - Steve's X Lock.

Some FM on the VFO on speech peaks caused the X Lock to drop out on TX. Dale W4OP fitted an X Lock to a 2000B a few years ago and mailed me a couple of mods he carried out. A solid state stabiliser for the 150V to the VFO (using the amazing LR8N3 chip and a TIP50) and simple FET buffer on the VFO output. The FET Buffer cured the FM, but I put the 150V regulator in as well.

It appears that KWs continued and rather baffling use of a valve VFO was a problem for every 2000 transceiver they made. The TX mixer 'pulls' the VFO slightly on transmit, causing it to move frequency a little - and opening the door for a bit of FM !. KW attempted to fix this on the B and E with a small trimpot on the IRT line that was only in circuit on transmit - they recommended listening to the VFO on another gen coverage RX and switching between TX and RX, twiddling it until there was no change in frequency. They FET buffer below appeared to solve this problem on my B and the trimpot could be shorted out.


The dreaded Kokusai 455KHz mechanical filter is pretty good, but as with my other KW gear the filter 'drifts' towards the LSB carrier crystal making USB a bit thin sounding. I have some ceramic resonators which make good carrier oscillators than can be moved slightly to solve this issue...The filter drifts because the mechanical element gets clogged with the deteriorating packing foam inside the can (over a period of 45 years it turns to a brown goo!). Cleaning the internal element of these filters is difficult (but not impossible) and normally brings them back to the correct frequency. All in all this radio appears in very good order.


Re-painting can be very daunting, but I enjoy it. For Drake and Heathkit I always use Hammerite - Drake is black and Heathkit is dark green. I am not too much of a purest for the exact colour - you'll never get a perfect match. All my Drakes look as new in satin black, and my little Heathkit looks great in satin dark green. Scratched cases and damaged bezels look dreadful and are easy to bring back to life, only a few days work...come on guys tatty radios look...tatty, give them a proper new life !

KW gear is a little harder as the cabinet and bezel are different colours (like Collins) - the main case colour is close to Ford Polar Grey, the bezel is lighter and Ford Dove Grey is a very good match - I defy anyone to tell the difference!.  I rub down with a light emery paper block soaked in water...and then use 3 light coats of car (automobile) grey primer and then 3 light coats of the top colour..always looks fantastic.

The completed cabinet below shows how good these two colours look

On both my 2000A and 2000B I have inserted an inline phono plug and socket in the speaker lead to the PSU. This allows the use of other speakers with the radio, or in the case of my 2000B cured a ground loop hum by plugging the speaker into the headphone socket rather than going through the rear multiway plug - strange but true!


Ebay purchase delivered in  Jan 2016 and I was more excited than a kid at Christmas. Opening the boxes revealed a near mint KW2000A with hardly any marks to show its age and the same with the Vespa, The 2000 had been well looked after and had been the subject of a part restoration. I sat stunned in front of them on the kitchen table. Now the hard part. The KW2000A PSU looked correct but did not appear to match the transceiver - on power up all the voltages checked out, but the radio just gently hissed. On close inspection two tubes appeared to have no heaters, the VFO and could this be?, surely all the heaters were wired together. Emails to KW collector and guru Guy G0UKN followed..he told me those two tubes have a separate heater supply...what??.  I had been fault finding with an early Schematic for the KW2000 (not the A,B or E) and this didn't show it. Then the penny started to drop - the Power Supply was for the earlier version of the radio and it didn't have the separate heater supply for the VFO. I traced the wiring and sure enough that was the problem, KW used a 9R dropper resistor in the PSU to get the 12V down to 6v for these two valves. A rummage in the junk box and the radio woke up!

Why the earlier restorer (who had done so much to the radio) missed this, I will never know. Perhaps he too was working from the wrong schematic and just couldn't work it out. Any how back to life and good power out on 40m, 80m amazing !!

Next I turned my attention to the KW Vespa Transmitter and its obviously re-built power supply, oh dear, what a mess. No way was I turning it on in this state - even more dangerous than when it came out of the factory in its odd cardboard cover, re build required. I changed caps, fixed awful dry joints, corroded fuse holders and put decent HV diodes in for the 750v HT..the ones in there did not look all that great.

After a 2 day total rebuild and test, the PSU was in good working order. Firing up the radio was met with ZERO output..not a milliwatt. Out with the scope and it soon became clear that the dreaded 455KHz Kokusai mechanical filter was as dead as a dodo, 455KHz carrier went in and nothing came out on the scope As a quick and dirty check I bypassed the filter with 1000pf...Bingo !!! Double Sideband and loads of drive. In the box of bits that came with my purchase there was 2 other filters one opened and one sealed..On a whim I put the sealed one in the VESPA and it came to life again...super reports on 40m with a Shure 444. Later I found some resistors had gone high, and the curse of the KW Owners manual caught me out several times. Its one area where they really cut corners - the manuals are awful, child like drawings, schematics and charts with errors. Compare them to Swan, Drake and Collins and you will see what I mean.


Next purchase a very clean KW201 Receiver from the 2016 Canvey Rally (Essex UK). Cheaper than eBay and promised as working on AM and with a good Mechanical Filter - but not good on SSB. This was a quick fault find quickly traced to an almost shorted coupling cap from the anode of the product detector to the grid of the Audio O/P valve. That's why it worked on AM but not SSB. This receiver has a 3KHz filter which is too narrow for AM but means that SSB sounds great provided there are no adjacent stations - of course you can add the optional Q Multiplier - but I have never seen one for sale  BEWARE the old carbon composition resistors in these KW gear, in the KW201 2 failed within a few days of getting it going, they all go high, but some just go VERY high.. Although the 201 came out at the same time as the VESPA transmitter they cannot 'transceive' together and have to be manually netted together - this an another odd KW design decision, even the later KW202 and 204 'separates' had to be netted onto the same frequency and they were in production into the mid 70s.  Drake separates all allowed proper transceive operation.

The Vespa and KW 201 are really only the TX and RX parts of the KW2000 transceivers and had similar limitations - only one filter bandwidth meant poor AM and CW performance, 200KHz band segments meant part coverage of 15 and 10m and did I say no transceive !


With the next  KW purchase it dawned on me that I had collected every item in this 1968 advert from the RSGB RadCom magazine from 1968. The prices were quite revealing as well - I had  paid less in  2016 for each item than the 1968 price. The only trivial difference was that my KW Vespa was a Mk1 and not a Mk2 model (that had a sweep tube PA). The KW1000 was another stunning eBay purchase from ham about 60 miles away. The amplifier is in A1 cosmetic condition - which is very rare for any equipment over 40 years old. In true KW tradition the mechanical construction can best be describes as agricultural. There are no EHT  interlocks and once removed from the case there is no cover over the PSU and the HV capacitors are facing not a place to put your fingers. All the panels are held together with different screws, but at least its rugged!.

Initial physical checks revealed a number of things that would need attention. The most obvious was the that the anode RF choke was burnt in the very middle, this a classic symptom of a resonance that has resulted in severe RF heating in the choke. After chatting with other owners of this linear it looks as if either a different choke was fitted at the factory or it was changed at some point, the output RF choke was also missing and not knowing why I decided to change the HV 'doorknob' capacitor as well. All the HV caps tested OK although the balancing resistors are quite low value at quite a bit of heat. I'll make up a PCB with modern caps and higher resistors at some point. My next mod before switch on was to insert an 800mA fuse in the EHT in series with a 10R 'glitch' resistor to limit the current should there be a tube flashover. I wound 2 different anode RF chokes - the first was close to the Heathkit SB200 design of 50 - 60uH, this worked OK but I was concerned that the inductance was too low for 80m and there could be RF on the EHT smoothing caps (despite extensive decoupling), in the end I went with a value over 120uH and a winding gap in the middle. I couldn't tell much difference but the engineer in me feels better !

Pictures of new choke and the burnt one on the right. Good power out on 40 and 80 from the Chinese 572b tubes already in the amplifier...over 500W out when driven from my KW2000A


This Pic of the KW1000 linear shows the plastic sleeved 800mA microwave oven fast blow fuse, I have also added a high power 10R 'glitch'  resistor to limit the current should the tubes ever flash over..not a problem years ago but Chinese 572 tubes are notorious  for variable quality and cannot really be trusted.  Those 25K balancing resistors sure do get hot..modern caps and higher value resistors to add soon











Next KW item added to the inventory a very nice KW Atlanta transceiver with matching external VFO, PSU and Shure 444 microphone. This has been well looked after and is in good working order. At some point in its past life its been modified to use 6146 tubes in place of the 6LQ6 'sweep tubes', this appears to have been well done and the power out on 40 and 80m is around 100w. There is very little restoration work to do, although as with all older kit I will change the PSU caps for more modern ones - and for some reason the HT choke in the PSU is missing. There are very few Atlantas on the air in the UK and little information on the web about them..initially aimed at the US market, KW did sell them in the UK from 1969.

The KW Atlanta is pretty much a direct component level rip-off of the USA Swan 500 ( I have checked the schematics!). OK so they look a 'bit' different, but the radio architecture, tube lineup, audio derived AGC, dial arrangement are all identical..I have no idea how KW got away with copying a US rig, knocking it up in the UK and selling it back in the US!.  KW copied the PSU as well but made it cheaper (surprise), the smoothing capacitors are all held together with cable ties and one bolt. The carrier 'bal' control is a nightmare in 'tune' mode - all the way left does not mean OFF, thats in the middle..turn all the way to the left and your back to lots of anode current...caught me out

The Swan 500 manual is a delight and I could probably use it to fault find the Atlanta, pity KW didn't copy that as well !. To cut cost the Swan design doesn't have any crystals as a premix to a 1st IF, instead the VFO frequency is switched by a bank of caps - of course KW copied this as well. Its a cheap 'Single Conversion' design with drift and poor dial calibration on 10m, lots of VFO trimmers, and the pi tank is used on receive as well as TX. Having said all that its fine on the low bands and the audio is very good if you back off the RF again to stop the audio derived AGC pops. Now I know why I love my Drakes so much. I have included a pic of the Swan 500..same radio but better built.

In Jan 2017 I noticed a problem with the Atlanta that had probably been getting worse since I obtained it..the S meter would gradually move upscale as the rig warmed up - as if the RF Gain was being backed off.  A blip of the PTT put the S meter back to near zero (if no signal was present) and then it would gradually move upscale over 15 mins. This was clearly an AGC fault as the rest of the radio was fine. Before pulling the radio apart I searched for AGC faults on the SWAN 350/500 and sure enough Chris GM3WOJ (also an avid KW collector) had come across the same problem in a SWAN. There is some sort of diode clamp in the AGC circuit (not sure what it does - there is no info in the manual)..but if the diode leaks a tiny bit it gradually deafens the receiver. It took me a few mins to find it (no layout diagrams exist), but I changed it for a 1N400x to see what would happen...problem solved


Loads of Hams in my early years bought a KW'EZEE' Match, it was the only easily available balanced tuner without a lossy balun. Power handling was bit limited, 100 - 200 W at the most. It could arc over in some matching situations with too much power. Still sought after - although some just ask too much money for them.  KW103 Power / SWR meter has been in my possession since the 70's. KW did put the match and the meter into one box (KW107/9), but people ask silly money for them (more than I have paid for a complete transceiver in some cases). After bartering unsuccessfully with a trader for a repainted KW101 SWR meter (he wanted £20 I offered £15) at the Eastbourne Rally in 2016 - he told me 'its KW mate and collectable'. Funny thing is one went on eBay for £10 a few days before, I gave up and walked away..same rally £20 for a kokusai filter, trader would not take £10. I paid less on eBay. Get real guys - selling old ham radio stuff is a hobby - not a business


Rare KW105 - Bought from a friend in 2017 the 105 does not appear to have been made for very long in the early 70's and was quickly superceded by the 107 and the higher power handling 109. Its still and 'EZEE Match' inside with unbalanced connections for the 80 - 10m bands on 5 separate sockets, and the two sets of balanced connectors which are not switched and must not be used at the same time as the unbalanced ones. The 105 switch positions are not marked on the front and this must have been confusing as they are marked on the back !. There is no 'through' position meaning its in circuit all the time and the SWR meter is not calibrated as a power meter as it is on the later models. Its super clean unlike many of the expensive 107s I see at rallies in the UK.