Clifton Laboratories 7236 Clifton Road  Clifton VA 20124 tel: (703) 830 0368 fax: (703) 830 0711



To search within the Clifton Laboratories site, enter your search term below.


Software Updates
Softrock Lite 6.2
Adventures in Electronics and Radio
Elecraft K2 and K3 Transceivers



December 2006 Archive


30 December 2006

Stan, W5EWA, measured his K2 using the procedure described at K2 RX Sensitivity. I've posted his measurements at that page along with mine to serve as reference points for anyone wishing to compare their K2. I'll also post any similar data provided to me by readers.

29 December 2006

I recently explained to another ham with a newly acquired signal generator how to measure the sensitivity of a receiver and I thought an expanded version of our discussion might be of general interest. I've accordingly added a new page K2 RX Sensitivity with the measurement methodology and results for my K2 transceiver.

28 December 2006

I received a couple of CD-ROMs from Mike, W2PY, full of high quality photographs of his Z90 assembly. I've posted one of the Z90 in his shack at User Views page, and also added several photos showing how Mike installed the Z10010 bandpass filter inside his Z90 enclosure, as first suggested by Doug, N6TQS, to the Construction Updates page.

Mike's assembly is some of the best work I've seen, by the way, although the limited size of the photographs I'll post won't let you see the details.

I had a telephone call yesterday from a prospective Z90 purchaser interested in using it with something other than a K2. I would appreciate hearing from any Z90 user who has installed a Z10000-U universal buffer board in equipment other than the K2. All my receivers (other than my K2) have IF-out ports so my testing at 40.455 MHz or 8.8 MHz simply involves plugging the Z90 into the IF Out port of my receiver.

24 December 2006

I've been remiss in not updating the site, but the holidays have made things a bit hectic around Clifton Laboratories.

Matt, KC0UKK has posted a review of the Z90 at Thank you Matt and I appreciate the nice things said in the review. If you are a Z90 or Z91 owner and care to add your views to Matt's please do so.

I have a sort-of update on the replacement panels. On 28 November 2006, Emachineshop's order status said the replacement panels were scheduled for 31 Dec 2006. As of today, the scheduled completion date is 04 January 2007.

Also on the update front, I've mentioned on this Updates page the random character problem experienced by one Z90 customer that I was able to duplicate here, but unable to diagnose other than some subtle incompatibility in Crystalfontz's new RoHS controller board. I sent a Z90 and sample displays showing the problem to Crystalfontz in early December for their analysis, but have yet to receive a response. I imagine the holidays have slowed their work as well.

I've also mentioned the next Clifton Laboratories project, and I've made progress on it as well over the last week. I'll call this project the "Z200 ECHO" for the moment. It's a small box with a 64x128 graphic display intended as an accessory to your K2 transceiver. The Z200 is still highly experimental and is a kludged breadboard state, but I hope to lay out a PCB for a prototype before the end of January.

It performs several tasks, as seen in the following photos.


Z200 in "Echo" mode -- it displays the K2's operating parameters, including several not visible without pushing a button on the K2. The vertical bar on the right is an S-meter and  tracks the K2's bar graph. The RX/TX display shifts, depending on the split mode.

The less-obvious parameters are attenuator/preamp status, AGC speed, audio filter status, RIT value and mode.

The Z200's audio spectrum analyzer function, principally to help in zero beating CW mode. This is a fast Fourier transform (FFT) based display and updates about twice a second. My K2 is set for a CW BFO pitch of 600 Hz, and as the photo shows, I've tuned in a signal at 600 Hz offset.

The large set of bars at the left are the so-called DC leakage and are inherent in FFT.

The FFT settings are 0-2 KHz, with 32.5 Hz bins, although options are available for a 4 KHz and 8 KHz span, with proportionally wider bins.

Oscilloscope mode.. Maximum useful frequency response is around 3 to 4KHz. The photo shows a 600 Hz beat  note against WWV obtained from the K2's audio in CW mode. The "jaggies" are visible in this image because it was taken with the Z200 prototype operating in reduced sample rate mode.
I've considered and breadboarded up a partial version that includes remote control functionality, permitting you to tune the K2, adjust settings, etc.  However, to do these functions right adds a lot of cost, as a good rotary encoder is necessary. In addition, the enclosure must be larger, which also adds to the cost. And, there are some issues surrounding the K2's interaction over the serial port that make remote control less than fully responsive in some circumstances.

The Z200 connects to the K2 via the K2's serial output port. It also requires an audio connection to perform the FFT and oscilloscope functions, and 12V power at approximately 100mA.

At the moment, the Z200 is a work-in-progress and the direction it takes has some flexibility. What do you think? If you have time, let me have your views on the following questions:

  1. Is the Z200 something you would find useful?
  2. Would you purchase it as a kit, with a tentative selling price of $150? All through-hole parts, no surface mount.
  3. Would you rather have it configured as a more fully functioned remote control unit (call it a Z300), allowing you to change frequency, mode, etc., from controls on the Z300? Would this extra functionality be worth an extra $100-125 in price to reflect the added complexity and increased enclosure size?
  4. What other things would you like to see displayed?
  5. Is it important that the Z200 have both an input and output RS232 port so that you may "daisy-chain" it with a PC running one of the various K2 control programs?


15 December 2006

Roger, K8RS, reports that his Z90 is playing and I've added a photo at the User Experience page. I've had a total of three bad parts in the Z90 deliveries to date, and Roger had the bad luck to have two of the three.

I had an enjoyable visit today from Mike, W4XN, to discuss the care and feeding of Vector Network Analyzers. Mike brought his HP8505 and we compared measurements between it and my HP8752B and went through a variety of measurement techniques.

Mike, W4XN, standing next to his HP8505 VNA, in my basement workshop. The 8505 is in four separate modules, from top to bottom being the storage normalizer, the display unit (CRT and receivers), the signal generator unit and an S-parameter test set (directional coupler and coaxial switch).
13 December 2006

I received my first Z90 repair Monday and I found an interesting problem. The 40-pin socket had several poor spring fingers to the point where the PIC's pin would float in the center of the fingers without making contact. In this case, the fix was to gently reform the damaged spring fingers, but the socket will likely require replacement in the future, as this is a temporary fix at best. After making this fix, the Z90 worked perfectly, so the builder had made no errors in assembly.

I don't know why the contact fingers failed, but it's worth inspecting your socket before installing it. In this case, the builder's kit passed the stage tests, including one that required measuring the ADC reference voltage at Pin 5, which was one of the faulty socket pins. I believe the pressure of the voltmeter probe made contact and thus the stage check was passed, but as soon as the probe was removed the connection failed. I zeroed in on missing ADC Reference voltage quickly as the likely source of the problem, but I must admit to scratching my head a few times as to why I could measure the correct voltage at a couple of test points but not on the PIC's pin, particularly when the voltage was correct with the PIC removed from the socket.

I also received Z90 completion comments from Mike, W2PY, that I've posted at the User's Views page. Mike had a slight problem with the Z90-Control updates, and I have decided to also place a complete Z90-Control package on the Software Update page, not just the updated files. 



10 December 2006

I've received construction completion reports from Randy, K4YYX, and Mike, W1BNC, and have posted their comments at the User Views page.

I had a request to make some short duration RF power measurements today, and I thought the K2 operators might find the keying waveforms interesting. The settings are 50 WPM, internal keyer, 3.7 MHz, 100 watts commanded output. The waveforms are captured with a Tektronix TDS 430A digital oscilloscope, in envelope mode, connected to the K2's output through a Bird 8323 100 watt, 30 dB attenuator and a Minicircuits UNAT-6 6 dB attenuator. The TDS 430's input impedance is switched to 50 ohms for these measurements.


50 WPM single dot
50 WPM, single dash
The RF output waveform has a rise time of about 5 ms and exhibits no overshoot. The decay is a bit faster. The resulting waveform sounds pleasant to the ear with no trace of click in my quick observation.

I commanded the K2's output power to 100 watts, and measured 1.097V RMS at the output of the attenuator chain, corresponding to 96.3 watts, excellent agreement with the commanded power. This computation assumes the attenuators are exactly 36 dB and neglects the connecting cable and adapter loss. My K2 was built by Frank, W6NEK, and I believe he set the K2's power control parameters using an LP-100 wattmeter from N8LP. I don't know if Frank had Larry calibrate his LP-100 or if he calibrated it himself, but in either case, my K2 agrees quite closely with my best power measuring ability.

The other thing I've been working on over the last few days is the touch screen input for graphic displays. I have a crude read/write program running with a 320x240 screen and a Fujitsu 4-wire analog resistive touch screen overlay. Here's a view of the results.

While my handwriting is not the best, it's actually not that bad. However, I am still working on the calibration routine that maps X resistance and Y resistance readings to X,Y pixel coordinates, so the little squares would appear on the LCD a half-inch or so from the end of the stylus, making precision writing difficult.

I should mention all this is being done with an 18F4620 PIC with code written in Swordfish. If you are a PIC programmer, you owe it to yourself to check Swordfish out, as it is an excellent product at a very reasonable price. Swordfish is available directly from Mechanique in the UK, or from several U.S. distributors. And, a free "SE" version is available for evaluation from the link.

04 December 2006

Walt, KE8BQ, finished building his Z90, and I've added his comments to the User Views page.  If you have finished your Z90 and would like to have a photo of it sitting next to your K2 displayed on the User Views page, send me an electronic image and I'll add it!

I was asked today whether I received a check, mailed around November 8th for a Z90. I responded to the purchaser with a "no," but something odd seems to have happened around that time, as this is now the second check that was mailed to me on the 8th or 9th that never arrived. When I receive a check, I immediately (same day, usually within an hour or two) respond with an E-mail message confirming that payment has been received and supply an estimated shipping date. So far, the shipping date has been the next day the post office is open, or at worst, one day beyond that. If anyone has mailed a check to me and has not received confirmation, please contact me. When I have the details of the mailing date confirmed, I'll visit the Post Office and ask if these failed deliveries should be investigated.

02 December 2006 AM

I've been tinkering with the CX display and experimented with a "dot" display mode. The dot is a 4-pixel square symbol placed on the screen -- no line. Here's a sample.

Sample "dot mode" display of a signal generator signal.

If there's an overwhelming demand, I can add dot mode to the current Z90 firmware as an option (line or dot) settable via the option setup menu.
01 December 2006

As usual, November's postings have been moved to the November 2006 archive page, accessible by clicking here or via the link at the top of this page.

I've worked more today with a Crystafontz "CX" series graphic display, with the S1D13700 controller chip. I have a version of the current Z90 software running as of this evening, although it needs further tweaking. The CX controller is available with the same blue/white CCFL negative display used in the current Z90. The two CX samples I have, however, use a transreflectance positive display.


A sample image of Z90 running on a transreflectance "CX" series display