Elecraft K2

Bullet-0 Home
Bullet-1 Auburn CA
Bullet-1 Andrea's Pages
Bullet-1 Fred's Pages
Bullet-2 Pick Your Nerd
Bullet-2 Amateur Radio
Bullet-3 General Info
Bullet-3 Contests
Bullet-3 Clubs
Bullet-3 My Station
Bullet-4 Elecraft K2
Bullet-4 Elecraft KX1
Bullet-4 Elecraft W1
Bullet-4 Elecraft Hexkey
Bullet-4 K6DGW/MM
Bullet-4 FOBB
Bullet-2 Homebrewing
Bullet-2 Other Stuff
Bullet-1 Events
Bullet-1 Sweep Riders

The Radio

My K2 is Serial Number 4398, and I purchased it in August 2004, which gives you some idea as to how many K2s there are out there! I have no commercial connection with Elecraft, other than being their customer. That said, you will find that I really like this radio. Its receiver has got to be in the top 5 in terms of blocking dynamic range, internal noise, sensitivity, and similar specifications. Signals just pop up from the background. It is simple to use, weighs 3.3 lbs, and is 3.4"H x 7.9"W x 9.9"D. I run it off of either a 12Ah or 4.5Ah 12V gel cell. It has an internal memory keyer, so the only accessories you need for a working station is a keyer paddle.

Mine is a "vanilla" K2 as I wasn't sure what options I really wanted when I purchased it. It operates 80m - 10m, up to 10W output, CW-only. I will undoubtedly install the RS-232 CAT interface and the amplifier keying modification so I can use it in contests. If I decide to make it my back-up rig for the annual CQP Alpine County Expedition, I'll add SSB and the KPA100 amplifier.

The Kit

The kit arrives in a box, in which are the chassis parts, several bags of parts and a manual. There was a bag for the control board and for the front panel board. The main RF board had two bags, there was a Misc parts bag, and a bag with some wire. A number of the parts (like crystals) are sub-packed in small paper envelopes, and everything is well labeled.

The first step before doing anything else is to inventory all of the parts against the illustrated parts lists in the back of the manual. While this is a bit tedious (there are lots of tiny parts), it assures that your kit is complete, and, possibly more important, allows you to identify all of the parts (e.g. some of the RF chokes look and are coded suspiciously like resistors). I separated the parts into small dishes on the workbench.

One very cool thing Elecraft did is supply all of the resistors taped and in the order you will install them. Eric and Wayne cannot know how much this helped me since I am totally colorblind. I checked each resistor with the ohmmeter before installing it, but I didn't have to hunt for the right one! Most of the capacitors are of the epoxy monolithic variety and have printed values (very small printed values!). I read through the manual and saw that the first capacitors didn't go in until a way into the project, so I let them sit in a bunch to start with. My wife tagged the RF chokes and other color coded parts with tape for me.


Required tools are fairly minimal but the correct ones are critical. An assortment of small screwdrivers is really nice although you can get by with only a medium sized phillips and flat-blade. Your wire cutters need to be good quality (Imports such as you might find at Harbor Freight aren't going to do the job well at all). Elecraft recommends flush-cut models which have the jaws beveled on the inside only. I have a good pair of standard diagonal cutters and they worked just fine. My mother-in-law was nurse many years ago, and as a result, I have acquired a few surgical tools that clamp, grip, and will lock in position. You don't need these, but if you can obtain one or two in a couple of sizes, they are fantastic for holding small parts, grabbing a toroid wire poking through the hole and pulling it tight, and the like.

WES51.jpgSoldering equipment is critical. You will need a good, temperature-controlled soldering station ... but then, if you plan on doing any electronics work, you'll need this anyway. I use a Weller WES51 set at about 670 deg which works great. Most have the sponge which you wet and use it to keep the tip clean. The tip of the iron needs to be small enough to get directly onto the solder pads and against the component lead, but also big enough to conduct an adequate amount of heat to the pad and lead quickly. This is important since heat traveling up the lead can damage some of the components if you have to leave the iron on the joint too long. My tip is about 3mm wide and 1.5mm thick and worked just great. Be sure and keep the tip clean of excess solder. The component holes on the circuit boards are plated through, and you'll find that placing the tip on the pad against the lead, and then touching the solder to the pad will immediately melt it into the hole. You want to fill the hole but you want a minimal amount of excess solder in the "fillet" on the lead.

Do not skimp on solder! Elecraft recommends 60/40, which is best. I used a 63/37 solder from previous projects. Mine is extremely fine (0.015"), like #28 or #30 wire. It gave me great control over just how much solder went into each joint, and I was able to fill the hole, bond to the wire, and not create excessive fillets. I would recommend the finest solder you can obtain. Really fine gauges do have the disadvantage that once you've pulled a piece off the roll, it will tend to get lost on your workbench, but you can cultivate the habit of always putting the solder down in the same place.

Magnifyer.jpg Be sure your work bench is well lighted and free of major shadows. Many of the parts are truly small, and can get easily lost if the light isn't good. And, unless you have the eyesight of a young eagle, you will welcome a magnifying light like the one in the photo on the right. The circular lamp around the magnifying glass removes nearly all the shadows, and makes soldering much easier. It also makes parts identification much easier by clearly illuminating the teeny weeny little guys and magnifying at the same time. For some of the monolithic capacitors, I needed a second hand-held glass under the light ... but then, I'm 64 and your experience may vary. I'd recommend you put a post-it note or some sort of book mark at the beginning of the parts list in the manual so you can get to it quickly. Some of the parts values will be hard to read, and possibly a little ambiguous at first. Referring back to the parts list will help you confirm that there are no capacitors with a value marking of "KR2", so turn the little guy over and read that number!


Here's my list of rules for assembly of all kits, but especially Elecraft:

  1. Do not work on your kit when you are: A. Tired; B. Stressed by other things; C. Distracted by family or work; or D. If you've had a glass of wine or a beer with dinner.

  2. Follow a standard procedure for installation of parts: For example, mine is:

    a) Read the manual step and jot down the Reference Designator (e.g. C15), the part description (e.g. electrolytic capacitor), and the value (e.g. 10uf). b) Find the Reference Designator on the silk screen on the board. Components go on both sides of the boards, however Elecraft has deftly arranged that you work on one side for awhile and then the other and they're pretty clear when you should turn the board over. c) Find the component and double check it's value against your notes and the manual. d) Re-identify the Reference Designator on the board, and then recheck the component value (it's easy after awhile to begin to confuse the ref. desig. like "C15" with a value code like "15"). e) Insert the part either as directed by the silk screen mask (e.g. diodes or electrolytics), or, where it doesn't matter, in a direction that will make the value readable after the board is completed. You can't do this in every case, but you can in most. Then f) bend the leads on the other side of the board to hold it in place.

  3. Elecraft will have you install a group of components, and then solder them all at once. They've thought this out pretty well, and I'd advise following their instructions. Since unsoldered components are a common cause of problems, before you start soldering, count the number of leads to be soldered and jot it down. Then, as you solder each lead, count it. When you "think" you're done, check and see if you counted the same number as you wrote down. If not, you're NOT done ... go find the one you missed and fix it.


You can buy pre-wound toroids from Mychael (aka "The Toroid Guy"), and if you're worried about getting them right, it's a good way to go. I wanted the "full experience" so I wound my own. First off, it is not hard, but it does require a little concentration. The worst one, wound with #30 wire is pre-wound for you from Elecraft. All the others are wound with civilized wire size (actually only one size), and have only a few turns. The pictures in the manual are correct, but you have to be very certain you are holding everything just like it shows in the picture. If your toroid is wound exactly like the picture, it will exactly fit into the holes in the board.

I'm told Eric is left-handed, and so am I. I don't know about Wayne, but my first try on the first toroid ended up with the leads opposite from the holes in the board. I carefully rewound it, with the same results. It only had a few turns so I asked my right-handed wife to wind it, and it came out right. Moral: If you are left-handed, be very careful. Once I "broke the code," it was a snap. The winding directions on T5 are critical, and you need to be absolutely certain you have them right. Incidentally, Elecraft provides you with more than enough enameled wire for several "do-overs." In fact, the lengths specified in the manual are all longer than required, so you will have plenty of wire.

Preparing the leads on a toroid occupies quite a few words in the manual, and is a common subject on the Elecraft E-mail Reflector. It isn't nearly as hard as all the discussion might lead one to think, however. First off, the enameled wire supplied in the kit is heat-strippable meaning that it will sort of melt and then burn up if heat is applied. As a result, I would strongly suggest against trying to use anything like a knife to scrape the enamel off as you are almost assured of knicking the wire, and it is very hard to get the wire really clean down to the copper. Ditto with sandpaper, although it solves the wire nicking problem.

The best way is the "solder blob" method. Do one lead of one toroid, and you'll master this technique immediately.

  1. Find a way to hold your wound toroid. I use a little desk clamp thingy that I won in a radio club raffle a hundred years ago. It has a heavy base, and a couple of alligator clips on articulated rods, and is perfect, but anything will do (see comment on surgical tools above).
  2. Position the toroid so the lead you're going to prepare is straight and either horizontal or slanting down a bit. Bend it out from the core so you can begin stripping up close to the core. (The single largest problem people have with toroids seems to be not stripping close enough. This allows enough unstripped wire to poke through the PCB, and then the solder doesn't really contact the wire).
  3. Hold the tip of your iron on the wire up close to the core and start feed solder onto it. You'll need a pretty good blob. Just keep the iron in one place, feed the solder slowly, and wait for the enamel to begin to bubble.
  4. When it is bubbling, it will turn very dark. As you're feeding solder slowly, begin to draw the iron out along the wire. The enamel will continue to bubble and disappear as you do this and the wire behind it will be tinned perfectly. You must keep feeding new solder however since that provides a fresh supply of flux, and the burned enamel will slowly contaminate your blob.

It is harder to explain than to do. I usually prepare slightly longer leads than I am going to actually want, and then clip off the ends since there is often a little mini-blob that hangs on there and prevents the wire from going through the hole. Also, be sure you have long enough leads that once installed, you can grab them with a tool and pull them tight.


I banned the vacuum from my shack during the assembly. I really recommend this if your shack/work area is prone to cleanliness activities. I really had only one problem during assembly and alignment. I got to the first alignment step, and all seemed to go well, until I tried the CAL PLL process. I kept getting INFO 232 which the manual said meant I wasn't on 40m (the only band that worked at that point). I was on 40m, and I'll spare you the details, but it turned out Elecraft had received a batch of transistors for Q9 and Q10 (the preamp for the frequency counter) that had low (or no) gain at higher frequencies. This caused the frequency counter to count OK when aligning the early steps (lower frequencies), but fail at the higher frequencies involved in the CAL PLL alignment. When it did not report the proper frequency at TP1, the firmware decided the radio was not on 40m and wouldn't run the linearization process. Elecraft sent me new transistors from another manufacturer, and all was well. There is a Builder Alert regarding this problem on the Elecraft Web Site.

My S/N is 4398, and I think I may have been the first or nearly the first customer with this problem. Elecraft got it fixed fast, so if your S/N is in the later 44xx series or higher, I doubt you'll see this. The bad transistors are Motorola guys, with a bluish face (or so I'm told). If yours are all black, you're probably home free. If you have the blue ones, Elecraft will replace them as soon as you tell them.

I experienced a "technical difficulty in my mind" while troubleshooting the above, which really isn't an Elecraft problem. Being a former engineer, I take things pretty literally, and when the manual said (or implied) that at the first RF alignment steps, the 40m receiver would be operational, I took it at face value. My receiver was inoperative on 40m or any other band, however, and of course, I feared the worst. It turns out that until you have run one "CAL PLL" step successfully, the receiver is not operational, and until I replaced Q9 and Q10, I could not complete that step. Once I had, the radio came alive on 40m just fine


The K2 is not "Your Father's Radio." A significant amount of it's functionality is resident in, or controlled by, the firmware in the main controller. I won't pretend to describe Wayne and Eric's design in exquisite detail, even if I could, but I can offer some insights that may help you.

4MHz Ref. Oscillator:There is a crystal controlled 4MHz oscillator on the Control Board. It's purpose is to provide the time base for the Frequency Counter, and it's frequency is adjusted by C22. Ultimately, since you will use the frequency counter to align the radio, getting this oscillator on frequency is important. You might thus think that you can tune the radio to 10.000 MHz, and then adjust C22 to zero beat the WWV signal (I did). As I was, you will be disappointed, in that changing C22 has no effect on the receive frequency.

What really happens is that you set C22 to some value, and then use the frequency counter to run the CAL PLL alignment process (takes about 7-8 minutes). That process stores a table in the EEPROM which ultimately determines what the frequency display says when you are receiving a given frequency. You then check to see what the display reads when zero beat with WWV. If it isn't right, you adjust C22 a little, re-run the CAL PLL function and check again. It sounds difficult and time consuming, but it really isn't. I got mine within about 15 Hz against WWV with four iterations ...took about 45 - 60 minutes.

Now, if you have a very well calibrated external frequency counter, you can actually measure the 4MHz reference oscillator and set it using C22. There are a couple of alternative methods on the Elecraft Web Site too. The important point to remember is, "Adjusting C22 while listening to the receiver will change nothing."

RF Alignment: The RF alignment procedure is very straightforward in the manual, but must be accomplished in the exact order shown. For some bands, a single inductor is used with a capacitor switched in for the next lower band. The receiver and transmitter also share the tuned circuits. The manual has you align the receiver first, and then the re-tweak the tuned circuits for the transmitter. In my case, the transmitter was producing a little over rated power on all bands after I finished the receiver alignment, and I never did tweak anything for the transmitter.

AGC Threshold: The AGC Threshold setting determines the input signal level that will begin reducing the gain of the receiver. It is adjusted by R1 on the control board, and the initial recommended setting is for 3.80 VDC at pin 5 of U2. The goal is to have AGC action start fairly quickly as signal strength rises, but you do not want the internal noise in the receiver to be activating the AGC. I followed the recommendation and moved on. Long after I had finished the radio, I became curious, disconnected the antenna, turned the preamp on, and switched between AGC ON and OFF. Sure enough, when I turned AGC ON, there was a perceptible decrease in the (very weak) receiver noise indicating that the noise was just barely activating the AGC. A small tweak to R1 fixed this, although my DMM still indicated 3.80 VDC.

IF Filter Alignment: This process took me awhile, but once I began to understand what is going on, it started to make sense. The four CW crystal filter(s) all have the same lower cutoff frequency. As you select increasingly narrow filters, the upper skirt moves downward in steps. The goal, of course, is to have the signal centered in each passband with a constant sidetone regardless of which filter is selected, or whether or not the radio is in CW or CW REV (at least that was my goal!)

The K2 provides a BFO frequency for each filter, for both CW and CW REV, and the CAL FIL process is how you select and adjust them. Your ultimate choices are stored in EEPROM and are then selected depending on the filter and mode. (There is a set for USB and LSB as well).

I'd recommend following the steps in the manual and writing down the BFO frequencies as indicated (use pencil, they'll change!). I had a very hard time achieving my goal with the method in the manual, but that's where I finally figured out how this all works, so it's worth going through it. Next, if you have a computer with a sound card, you can download "Spectrogram" from www.visualizationsoftware.com/gram/gramdl.html. It's shareware, and it lets you actually see the filter passbands and where the desired sidetone will place the signal. There are a number of postings on the Elecraft site on using this method. Beware of one pitfall, however: I found that the switching power supply for my laptop put so much noise into the sound card that I couln't see the passbands!

In truth, I only used Spectrogram once. I found it hard to achieve my goal of a constant sidetone regardless of filter selection. That one usage was very helpful to me in visualizing and understanding how this all worked, however.

I ultimately viewed the CAL FIL process as being able to customize my radio to the way I wanted it to operate. Having gotten it close with the manual technique, I then used my service monitor as a generator and tweaked everything right on. It took an afternoon, but now, the signal is centered in the passbands when the CW pitch is equal to the sidetone frequency (600 Hz for me), for CW. Switching to CW REV nets a small change (60 Hz or so), but in that mode, subsequent filter changes still yield a constant beat note. I have never needed CW REV, so at this point, I declared a victory and shut the service monitor off. I left the LSB and USB settings where they came out using the process in the manual, which seems to work OK.

Power Setting: You'll see quite a bit on the reflector, in the tech notes on the Web Site, and in the reflector archives about the POWER control. Like a lot of the K2, it is again tied up in a control loop through the main processor. The control tells the processor the power you want (i.e. "requested power") The processor sets the ALC such that you get that power out of the radio. Works great, just not how you might expect. I think mine actually outputs about 5.5W when I set the power control to 5.0W.

So, How's K2 #4398?

Well, in three words, I love it. I developed an interest in contesting after I retired, and I tend to do that QRO. It is fun to bust a pileup on a DS2, and that usually takes some power. The K2 has really fostered an interest in QRP operation -- totally different than contesting, but lots of fun. The K2 receiver is noticably better than my Yaesu FT-847 on HF (not where I normally use that radio), and at least discernably better under some conditions than my Kenwood TS-850S. I like the size, weight, and functionality ... the knobs and switches you need are there and those you generally don't need very often are not there (they're in menu settings). I will be getting the RS-232C interface, and the ATU. I may get the SSB option even though SSB is a tough mode for me, and I might even get the 100W amplifier although the only time I'd need another 100W radio is as a backup rig for the annual CQP expedition to Alpine County. I will definitely not be getting the 160 option -- life is way too short at my age for QRP on 160.

The K2 is easy to build if you're willing to go slowly and get things right the first time (some of the parts install very easily but remove extremly hardly!) For me, parts identification was by far the most difficult, which might seem strange for someone with just over 50 years as a ham. However parts have changed over the years, and in 2004, RFC's look like resistors, capacitors are tiny little things with even smaller identifications, and I don't think binocular cores existed when I assembled my last Heathkit. (Rest assured, there are no wax impregnated paper capacitors in the K2!) I would recommend the K2 to anyone with some basic experience in electronics, electronic parts, and ham radio, and you end up with a world-class radio. It may be a bit daunting for a first-timer, but then there are other simpler Elecraft radios to get yout started.

Top of Page | K6DGW Home | Fred's Pages | Home

Copyright ©, 2004 All Rights Reserved/Revision 1.00 11/18/2004 1200Z
Page URL: http://www.foothill.net/~andreaj/Ham41.htm
Author: K6DGW@arrl.net

Permission is granted to reproduce any and all textual material in this HTML page for non-commercial purposes only. Reproduction or use of any graphic or photographic image is prohibited without the written permission of the site author and all persons appearing in such images. Reproduction and/or distribution of any software products downloaded from these pages is subject to the terms and conditions accompanying the downloaded files. All logos are the copyrighted property of the organization(s) represented, are used here with permission, and may be reproduced and/or used only with the permission of those organization(s).