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


October 2009 Archive


29 October 2009

The Z10000 buffer amplifier is now replaced with the Z10000B version.

The Z10000B's design and performance specifications are identical with the original Z10000. There is no price change with the "B" version. The main differences are:
  • Changed to a surface mount voltage regulator
  • Slightly smaller printed circuit board
  • Rearranged parts layout for easier building
  • Bottom board surface now has silk screening
  • Component ID now starts with R1, C1, etc.

Since the performance specifications are not changed, I have not revised references to the Z10000 to add the "B" suffix on this page and elsewhere on this web site. Hence, unless otherwise specifically mentioned, information on the Z10000 applies equally to the "B."

The photos below show the Z10000B (left) and the original Z10000 (right).

28 October 2009

I'm at the thinking and paper design phase of a new kit. I don't want to say much about it until the project is much closer to production, but I have a question for my readers and prospective customers�surface mount or through hole?

The new kit is not RF so I have some flexibility on parts choice and likely can design it as either a through-hole or surface mount kit. If surface mount, none of the ICs will have a pin spacing of less than 20 pins/inch, which is  the same pitch as the AD8007 in the Z10000 buffer amplifier. At a rough guess, the kit will have 20 to 25 ICs. Double sided printed circuit board, with solder mask and silk screen. If surface mount, there will probably be some components on the bottom surface. If so, the bottom of the board will also be silk screened. A through hole PCB will be 5" x 7" and the surface mount PCB would probably be 4" x 6." Since I have not even completed a full schematic, these are rough guesses.

Going surface mount probably means a slightly lower price, as the PCB can be smaller and some parts are a bit less expensive in surface mount form. At a total guess, surface mount might reduce the kit price by 5%. Performance should be identical in either through-hole or surface mount. If I go surface mount, the smallest passive parts will be 1206, which are the size used in the Z10000 buffer amplifier.

So, what about it guys (and gals if there are any reading)? Is it important to you that a kit be through hole? Personally, I find it easier to build a surface mount kit, at least the ones I design, but it's not a big deal to me one way or the other. In the event of a component failure, it's generally easier to replace a surface mount part and without risking damage to plated through holes.

I would appreciate your views - drop me an E-mail at the address at the top of this page.


24 October 2009

My sheet metal supplier says he plans to run the enclosures this coming week or the next. My fingers are crossed. If this schedule holds, I plan on shipping two or three active antenna kits and two or three distribution amplifiers for trial build in mid-November and if all goes well, quantity (well, quantity for me, at least) shipments soon thereafter. After the enclosures arrive, I have to build each kit, photographing the construction and write the instruction manual, which is at least a week's work if nothing intervenes.

Anyone interested in doing a trial build should drop me a note. I expect the trial builder to comment on the instructions, identifying errors and the like.

I'm still working on pricing for the active antenna and distribution amplifier but will have pricing finished before the trial build units ship.


23 October 2009

As I mentioned, my sheet metal supplier sent two of his standard design enclosures to me with the thought that it would be faster to modify a stock design than to set up for a completely custom design. Earlier today, I sent dimensioned drawings and photos to the supplier. I am not making promises on delivery date, but I'm cautiously optimistic.

The three photos below show the DC power injector, model Z1203. The Z1203 is used with the forthcoming active antenna and the Z10040B Norton amplifier when powered over the coaxial cable.

Of course, the final enclosures will have silk screened printed labels, not the hand written ones I added to my test layout!

The two photos below show the Z1300 antenna multi-coupler enclosure, with my markings, and some extra holes. As with the DC coupler, the Z1300 enclosure will have silk screen printing.

Both enclosures are 16 ga steel (around 0.060 inches thick), and are quite sturdy, to say the least. The surface finish is black powder coating, which looks very good and is extremely tough and resistant to scratching.


22 October 2009

I've added a new sub-page to the Z10040B Norton Amplifier page showing how 2N5109 transistors matched at a single point with the transistor gain function of an inexpensive multi-meter track over a range of base and collector currents as seen on a Tektronix semiconductor curve tracer. The page is 2N5109 Matching. The main Z10040B Norton amplifier page is Z10040A Norton Amplifier.

Yesterday I received the sample enclosures mentioned on 16 October. I've modified one for the DC power coupler today and will post a photo or two tomorrow after I build a coupler and install the parts. I'll work on the antenna multicoupler over the weekend and I should have the punch and silk screen drawings off to my supplier in the next two or three days.

16 October 2009

I've added a new page with forward current versus voltage for infrared, red, green, yellow and white LEDs. LED Vf vs If.

16 October 2009

I spoke with my sheet metal supplier yesterday and I believe we have a path for delivery of the parts holding up my active antenna and the multicoupler. Instead of the custom dimensioned enclosures he agreed to make, I'll go with his stock enclosures, with my hole punching and silk screening, of course. The stock enclosures are slightly larger than the ones I designed, but not so much that it's a concern. I should receive samples of the stock enclosures next week, and I'll mark them up for my designs.

I'm not making a prediction of delivery schedule for obvious reasons. I have a quantity of printed circuit boards for the active antenna and DC power coupler and a smaller quantity of antenna multcoupler boards, and the limiting factor remains the sheet metal work.

14 October 2009

I assembled a Z1300 antenna multicoupler and sent it to a beta tester today. The delay on releasing the Z1300 as a real kit is the enclosure, which is still delayed by my sheet metal supplier.

The photos below show the beta unit in a grossly oversize enclosure. The front/rear panels have an ink jet printed paper overlay with matte mylar laminate. The production enclosure will be black powder coated with white silk screen legends.

10 October 2009

I was asked how the capacitance function on a Fluke digital multi-meter compared with the dedicated GR 1658 and the HP 4192A LF Impedance meter.

I measured the same set of 1% capacitors this afternoon with a Fluke 189 digital multi-meter as well as an Agilent 34410A digital multi-meter. The Agilent data is smoothed by its averaging function.

  Fluke 189   Agilent 34410A (Avg Mode)
    Measured   Difference   Measured   Difference
0.22 uF 0.222 uF 0.91%   0.22181 uF 0.82%
0.01 uF 0.01002 uF 0.20%   0.009994 uF -0.06%
0.01 uF 0.010062 uF 0.62%   0.010057 uF 0.57%
1000 pF (A) 1000.7 pF 0.07%   (A) 999.276 pF -0.07%
100 pF  (A) 100.8 pF 0.80%   (A) 100.296 pF 0.30%

I did not mark the 1000 and 100pF capacitors so they are measured in random order in all measurements today and in the 06 October data. Hence, the most meaningful comparison for the 1000 and 100pF capacitors are the averages.

The Fluke 189 only displays capacitance to the nearest 1pF and has a rather large zero value, 68pF. Hence the 100pF capacitance values should be regarded with some caution.

    4192A-1658 4192A-189 4192A-34410A
Nominal   Divergence Divergence Divergence
0.22 uF -0.59% -0.73% -0.64%
0.01 uF -0.10% -0.36% -0.10%
0.01 uF -0.09% -0.17% -0.12%
1000 pF (A) -0.01% (A) -0.11%  (A) 0.03%
100 pF (A) 0.10% (A) -0.39% (A) 0.12%

In general, there's quite decent agreement amongst all the instruments, recognizing their rated accuracy and resolution:

Fluke 189:

Agilent 34410A:

09 October 2009

I'm an electronics and radio history buff, and recently ran across a fascinating documentary film on line. It's from the WW II era and shows how Halicrafters made the SCR-299 mobile radio station for the Army Signal Corps.

The SCR-299 may be more familiar to hams from the transmitter end, the famous BC-610, which was the amateur radio HT-4 transmitter modified for improved reliability and ruggedness. The film has amateur radio content, as well as an appearance by Bill Halligan, the founder of Hallicrafters. It's well worth watching to see how radios were made in the 1940's. (Bill Halligan's call was W9WGE at the time, which you will see at the outset of the film.)

The movie is in  two parts, each one a little under 15 minutes long.

09 October 2009

As an experiment suggested by Dave, G3TJP, I made a capacitor with vinyl food wrap dielectric. It uses the same chocolate candy foil plate as the earlier paper dielectric units. Details at  Roll your own Paper Capacitor which is a now slightly inaccurate title.

07 October 2009

I've added a new page showing how one might make a simple (but not very high quality) capacitor from a couple pieces of thin foil and paper. It's at Roll your own Paper Capacitor


06 October 2009

I mentioned acquiring a few 1% capacitors for a cross check on capacitance measuring devices I own. The two most accurate devices are an HP4192A LF Impedance Meter and a General Radio 1658 DigiBridge. The accuracy specifications for both instruments are a function of the component value, test frequency and oscillator level amongst others, but should be on the order of ±0.2% or better for the capacitors I measured and in most cases ±0.1%.

The results are below.

    HP 4192A LF Impedance   GR 1658 DigiBridge   4192A-1658
Nominal   Measured   Difference   Measured   Difference   Divergence
0.22 uF 0.2204 uF 0.18%   0.2217 uF 0.77%   -0.59%
0.01 uF 0.009984 uF -0.16%   0.009994 uF -0.06%   -0.10%
0.01 uF 0.010045 uF 0.45%   0.010054 uF 0.54%   -0.09%
1000 pF 1000.6 pF 0.06%   1003.4 pF 0.34%   -0.28%
1000 pF 1001.5 pF 0.15%   1000.9 pF 0.09%   0.06%
1000 pF 1004.4 pF 0.44%   1003.8 pF 0.38%   0.06%
1000 pF 995.0 pF -0.50%   994.4 pF -0.56%   0.06%
1000 pF 996.3 pF -0.37%   995.8 pF -0.42%   0.05%
1000 pF  (A) 999.56 pF -0.04%   (A) 999.66 pF -0.03%   -0.01%
100 pF 100.49 pF 0.49%   100.39 pF 0.39%   0.10%
100 pF 100.88 pF 0.88%   100.77 pF 0.77%   0.11%
100 pF 100.00 pF 0.00%   99.91 pF -0.09%   0.09%
100 pF 100.67 pF 0.67%   100.56 pF 0.56%   0.11%
100 pF 100.02 pF 0.02%   99.91 pF -0.09%   0.11%
100 pF  (A) 100.412 pF 0.41%   (A) 100.308 pF 0.31%   0.10%

The rows with measured value (A) are the average of the five individual capacitors measured.

From the data, a few observations can be made:

  1. All parts are within their ±1% tolerance.
  2. Except for two samples, the 4192A and 1658 measurements agree within the expected error tolerance. It's possible that with respect to the 0.01uF outlier that I accidentally measured the wrong sample with the 1658 DigiBridge. I'll revisit that at the next opportunity. The 0.22uF sample may exhibit some non-linearity with respect to the test voltage or some other factor, or it may reveal a range calibration issue. I'll look at it again as well.
  3. At 1000pF and below, measuring to the 0.1% level requires consideration of the lead capacitance and test fixture problem I discussed on 03 October. The 4192A calibration procedure corrects for fixture capacitance and the part can be slid into the fixture to remove most, if not all, lead-to-lead capacitance. The 1658 DigiBridge does not have an explicit open/short calibration process so the open circuit fixture capacitance must be subtracted from the reading. I've done that in the results tabulated above.


03 October 2009

I recently purchased a few units of several values of 1% tolerance capacitors from Mouser, 100pF, 1000pF, 10000pF and 0.22uF in particular. (I wanted 0.1uF but it was out of stock when I placed the order.)

In measuring the 100pF capacitor with an HP 4192A LF Impedance Analyzer, I ran into an interesting definitional issue�where does the lead stop and the capacitor begin?

My 4192A has a 16047A adapter which has connection fingers for each side. It's thus possible to insert the capacitor all the way into the adapter, as shown in the left photo below, or to allow it to protrude from the fixture as in the right image.

It is, or perhaps should be, obvious that allowing the capacitor to extend from the fixture adds the lead capacitance to the true capacitance. But, how much of a difference does it make to the measured value? After all, the leads are 0.2 inches (5.08mm) spaced and only about 0.3 inches (7.6mm) long.

I set the 4192A to 10 KHz and performed an open/short calibration of the instrument and fixture. With no capacitor installed, the 4192A reads a satisfying 0.00pF.

With the capacitor protruding from the fixture, as in the earlier righthand image the 4192A reads 100.02pF and inserted fully into the fixture 99.97pF, a difference of 0.05pF.


To see if 0.05pF is a reasonable value I first computed the capacitance of  two 0.015 inch diameter wires spaced 0.2 inches as 0.2 pF/inch. When fully extended, the leads are approximately 0.3 inches from the fixture, corresponding to 0.06pF, providing good agreement with the measured value.

There's probably a second order effect here, as each of the fixture clips see some capacitance to the other lead as well as the lead-to-lead capacitance.

03 October 2009

As usual, I've moved the September 2009 Updates to an archive page, readable by clicking here or via the table at the top of this page.